═══ 1. DEFINITIONS ═══ ═══ 1.1. name ═══ Term Description name Defect identification number ═══ 1.2. reference ═══ Term Description reference ═══ 1.3. state ═══ Term Description state Status of a defect. It could be OPEN,CLOSE,RETURN,VERIFY,CANCELLED or WORKING. OPEN: Defect has been reported, but it has not been worked on yet. CLOSE: Defect has been resolved. RETURN: Defect has been rejected for a specified reason. See notes. VERIFY: Defect has been worked on and it is waiting for test. CANCELLED: Defect has been cancelled. See notes. WORKING: Defect has been assigned to a developer who is working on it. ═══ 1.4. serverity ═══ Term Description severity 1-4 1 =>Very bad problem. 2 =>Bad problem. 3 =>Average problem. 4 =>Not important problem. ═══ 1.5. answer ═══ Term Description answer Answer Code ----- Type -------- Description asni answerAccept Assignment Incorrect asnm answerAccept Assignment Missing autoopen answerAccept DO NOT USE: Automatically opened b bldm answerAccept Build/Packaging/Merge Missing doci answerAccept Documentation Incorrect dsai answerAccept Datastructure/Algorithm Incorrect inti answerAccept Interface Incorrect lfi answerAccept Large function Incorrect timi answerAccept Timing/Serialization Incorrect lfm answerAccept Large function Missing timm answerAccept Timing/Serialization Missing intm answerAccept Interface Missing dasm answerAccept Datastructure/Algorithm Missing chkm answerAccept Checking Missing chki answerAccept Checking Incorrect bldi answerAccept Build/Packaging/Merge Incorrect docm answerAccept Documentation Missing asnip answerReturn PRS Assignment Incorrect - Perm Re asnit answerReturn TRS Assignment Incorrect - Temp Re asnmp answerReturn PRS Assignment Missing - Perm Rest asnmt answerReturn TRS Assignment Missing - Temp Rest autoopen answerReturn DO NOT USE: Automatically opened b bldip answerReturn PRS Build/Packaging/Merge Incorrec bldit answerReturn TRS Build/Packaging/Merge Incorrec bldmp answerReturn PRS Build/Packaging/Merge Missing bldmt answerReturn TRS Build/Packaging/Merge Missing chkip answerReturn PRS Checking Incorrect - Perm Rest chkit answerReturn TRS Checking Incorrect - Temp Rest chkmp answerReturn PRS Checking Missing - Perm Restri chkmt answerReturn TRS Checking Missing - Temp Restri docip answerReturn PRS Documentation Incorrect - Perm docit answerReturn TRS Documentation Incorrect - Temp docmp answerReturn PRS Documentation Missing - Perm R docmt answerReturn TRS Documentation Missing - Temp R dsaip answerReturn PRS Datastructure/Algorithm Incorr dsait answerReturn TRS Datastructure/Algorithm Incorr dsamp answerReturn PRS Datastructure/Algorithm Missin dsamt answerReturn TRS Datastructure/Algorithm Missin intip answerReturn PRS Interface Incorrect - Perm Res asnmp answerReturn PRS Assignment Missing - Perm Rest asnmt answerReturn TRS Assignment Missing - Temp Rest autoopen answerReturn DO NOT USE: Automatically opened b bldip answerReturn PRS Build/Packaging/Merge Incorrec bldit answerReturn TRS Build/Packaging/Merge Incorrec bldmp answerReturn PRS Build/Packaging/Merge Missing bldmt answerReturn TRS Build/Packaging/Merge Missing chkip answerReturn PRS Checking Incorrect - Perm Rest chkit answerReturn TRS Checking Incorrect - Temp Rest chkmp answerReturn PRS Checking Missing - Perm Restri chkmt answerReturn TRS Checking Missing - Temp Restri docip answerReturn PRS Documentation Incorrect - Perm docit answerReturn TRS Documentation Incorrect - Temp docmp answerReturn PRS Documentation Missing - Perm R docmt answerReturn TRS Documentation Missing - Temp R dsaip answerReturn PRS Datastructure/Algorithm Incorr dsait answerReturn TRS Datastructure/Algorithm Incorr dsamp answerReturn PRS Datastructure/Algorithm Missin dsamt answerReturn TRS Datastructure/Algorithm Missin intip answerReturn PRS Interface Incorrect - Perm Res intit answerReturn TRS Interface Incorrect - Temp Res intmp answerReturn PRS Interface Missing - Perm Restr intmt answerReturn TRS Interface Missing - Temp Restr invdld answerReturn Invalid - Down Level Driver invextbios answerReturn Invalid - External BIOS problem invexthdwe answerReturn Invalid - External hardware proble invextsw answerReturn Invalid - External software proble invibmbios answerReturn Invalid - IBM BIOS problem invibmhdwe answerReturn Invalid - IBM hardware problem invibmsw answerReturn Invalid - IBM software problem invinf answerReturn Invalid - Insufficient Information invsubcan answerReturn Invalid - Submitter Canceled invsug answerReturn Invalid - Suggested Enhancement invtcp answerReturn Invalid - Test Case Problem invunr answerReturn Invalid - Unreproducible invwad answerReturn Invalid - Working As Designed lfip answerReturn PRS Large Function Incorrect - Per lfit answerReturn TRS Large Function Incorrect - Tem lfmp answerReturn PRS Large Function Missing - Perm lfmt answerReturn TRS Large Function Missing - Temp timip answerReturn PRS Timing/Serialization Incorrect timit answerReturn TRS Timing/Serialization Incorrect timmp answerReturn PRS Timing/Serialization Missing - timmt answerReturn TRS Timing/Serialization Missing - valunr answerReturn Valid - Unreproducible asni answerAccept Assignment Incorrect asnm answerAccept Assignment Missing autoopen answerAccept DO NOT USE: Automatically opened b bldm answerAccept Build/Packaging/Merge Missing doci answerAccept Documentation Incorrect dsai answerAccept Datastructure/Algorithm Incorrect inti answerAccept Interface Incorrect lfi answerAccept Large function Incorrect timi answerAccept Timing/Serialization Incorrect lfm answerAccept Large function Missing timm answerAccept Timing/Serialization Missing intm answerAccept Interface Missing dsam answerAccept Datastructure/Algorithm Missing chkm answerAccept Checking Missing chki answerAccept Checking Incorrect bldi answerAccept Build/Packaging/Merge Incorrect docm answerAccept Documentation Missing asnip answerReturn PRS Assignment Incorrect - Perm Re asnit answerReturn TRS Assignment Incorrect - Temp Re asnmp answerReturn PRS Assignment Missing - Perm Rest asnmt answerReturn TRS Assignment Missing - Temp Rest autoopen answerReturn DO NOT USE: Automatically opened b bldip answerReturn PRS Build/Packaging/Merge Incorrec bldit answerReturn TRS Build/Packaging/Merge Incorrec bldmp answerReturn PRS Build/Packaging/Merge Missing bldmt answerReturn TRS Build/Packaging/Merge Missing chkip answerReturn PRS Checking Incorrect - Perm Rest chkit answerReturn TRS Checking Incorrect - Temp Rest chkmp answerReturn PRS Checking Missing - Perm Restri chkmt answerReturn TRS Checking Missing - Temp Restri docip answerReturn PRS Documentation Incorrect - Perm docit answerReturn TRS Documentation Incorrect - Temp docmp answerReturn PRS Documentation Missing - Perm R docmt answerReturn TRS Documentation Missing - Temp R dsaip answerReturn PRS Datastructure/Algorithm Incorr dsait answerReturn TRS Datastructure/Algorithm Incorr dsamp answerReturn PRS Datastructure/Algorithm Missin dsamt answerReturn TRS Datastructure/Algorithm Missin intip answerReturn PRS Interface Incorrect - Perm Res intit answerReturn TRS Interface Incorrect - Temp Res intmp answerReturn PRS Interface Missing - Perm Restr intmt answerReturn TRS Interface Missing - Temp Restr invdld answerReturn Invalid - Down Level Driver invextbios answerReturn Invalid - External BIOS problem invexthdwe answerReturn Invalid - External hardware proble invextsw answerReturn Invalid - External software proble invibmbios answerReturn Invalid - IBM BIOS problem invibmhdwe answerReturn Invalid - IBM hardware problem invibmsw answerReturn Invalid - IBM software problem invinf answerReturn Invalid - Insufficient Information invsubcan answerReturn Invalid - Submitter Canceled invsug answerReturn Invalid - Suggested Enhancement invtcp answerReturn Invalid - Test Case Problem docmt answerReturn TRS Documentation Missing - Temp R dsaip answerReturn PRS Datastructure/Algorithm Incorr dsait answerReturn TRS Datastructure/Algorithm Incorr dsamp answerReturn PRS Datastructure/Algorithm Missin dsamt answerReturn TRS Datastructure/Algorithm Missin intip answerReturn PRS Interface Incorrect - Perm Res intit answerReturn TRS Interface Incorrect - Temp Res intmp answerReturn PRS Interface Missing - Perm Restr intmt answerReturn TRS Interface Missing - Temp Restr invdld answerReturn Invalid - Down Level Driver invextbios answerReturn Invalid - External BIOS problem invexthdwe answerReturn Invalid - External hardware proble invextsw answerReturn Invalid - External software proble invibmbios answerReturn Invalid - IBM BIOS problem invibmhdwe answerReturn Invalid - IBM hardware problem invibmsw answerReturn Invalid - IBM software problem invinf answerReturn Invalid - Insufficient Information invsubcan answerReturn Invalid - Submitter Canceled invsug answerReturn Invalid - Suggested Enhancement invtcp answerReturn Invalid - Test Case Problem ═══ 1.6. symptom ═══ Term Description symptom appf Application Failure autoopen DO NOT USE! Automatically opened by CMVC. bb Build Break codb Cost of Doing Business (e.g. Layout) comp Compatibility dc Data Corruption fnwad Function not working as designed. prf Performance thipe Trap/Hang/IPE use Usability vis Visually Incorrect ═══ 1.7. tracks ═══ Term Description tracks ═══ 1.8. verify ═══ Term Description verify ═══ 2. DEFECTS ═══ Asteriks appearing at the beginning of the defect number indicate that the defect has been fixed in the 6.498 Beta ( 3/93 Beta ). ═══ 2.1. 24584 DESKTOP POPUP MENU: MNEMONIC S IS USED TWICE. ═══ name 24584 reference PM20/SM11781 abstract DESKTOP POPUP MENU: MNEMONIC S IS USED TWICE. duplicate state returned severity 3 answer valunr symptom use duplicate defects: duplicate features: none. notes: *------------------------------------------------------------------------------* * The following CMVC Note contains information that could not be directly * * mapped into a corresponding CMVC field. This data is directly from the IDSS * * PTRFV SUMMARY file for product: PM20 PTR number: SM11781 * * This Note was created on: 06/21/92 at: 11:52:51 * *------------------------------------------------------------------------------* Cross-Reference: POST - Reported Version: R203 Reference Number: - Type of fix verification: VERSION Test System: 307N - Documentation impact? Date PTR was OPENED: 04/13/92 Date PTR was ANSWERED: Date PTR was CLOSED: Date PTR was ASSIGNED to the current technical owner: 04/13/92 Date PTR entered its current STATUS: 04/13/92 Status PTR was in when it was converted to CMVC: OPEN The last date and time the PTR CHANGED: 06/17/92 at 07:45:25 Submitter data: L.L. LEVINE - LJLEVINE at BCRVMPCV Technical Owner data: L.T. BROWN - TREPANIE at ???????? Responsible Manager data: C. CALDWELL - CALDWELL at BCRVMPCV Submitter's Manager data: J.A. HUNT - JHUNT at BCRVMPC1 Version of the product this PTR was fixed against: Version status of this PTR: Date this PTR entered the above version status: Driver name this PTR was integrated into: Date this PTR was integrated: Date the fix was accepted for this PTR: >------------------------------------------------------------------------------< > The following CMVC Note(s) are from the original text of IDSS product: PM20 < > PTR number: PSM11781. These Notes were created on: 06/21/92 at: 11:53:04 < >------------------------------------------------------------------------------< >>>> PROBLEM DESCRIPTION <<<< The popup (context) menu for the desktop uses the "S" mnemonic for both "Create shadow" and "Sort". One of them needs to be changed to be a unique mnemonic for this popup. Would probably recommend changing sort mnemonic to "r". >>>> ANSWER TO PROBLEM <<<< >>>> END OF PROBLEM TEXT <<<< This is unreproducible in the latest drivers. -Sheila 2-XXXX ═══ 2.2. 46524 CRITERIA CHANGED ON PAGE 3 OF THE INCLUDE SETTINGS FOR A FOLDER ═══ name 46524 reference abstract CRITERIA CHANGED ON PAGE 3 OF THE INCLUDE SETTINGS FOR A FOLDER duplicate state verify severity 2 answer autoopen symptom autoopen duplicate defects: duplicate features: none. notes: CRITERIA CHANGED ON PAGE 3 OF THE INCLUDE SETTINGS FOR A FOLDER WHICH IS A SHADOW, IS NOT SAVED AFTER SHUTDOWN AND REBOOT. APAR# PJ04691 APAR created by DITZIG, LINDA D. Customer rep DAVID WILKINS Customer phone number 205/251/2985 5/1 1) Create a folder on the desktop. 2) Drag from one of the drives a directory containing .exe file and other files, to this folder as a shadow (CTRL-SHIFT-BTN2 3) From the Include Page 3 of the settings notebook,Add criteri to the shadow folder's settings-select ADD pushbutton and select: Property to be compared - Real Name Comparison Type - Not Equal Comparison File - *.EXE Use of Criteria - AND Press the ADD pushbutton again 4) Close the settings notebook and the shadow folder will no longer display .exe files. This should work. 5) Now shutdown and reboot, look at your shadow folder again an the .exe files are back! This is easily recreated on HPFS and FAT. PMR#: 8X542,091,000 -ADAMS, MARTHA - -L148/OS2SL -P3S3-92/05/12-10:31 -CE LV 2 CB/SEV 3 PHONE GOOD CUSTOMER IS HAVING PROBLEMS WITH SETTING WINDOWS. -ALLEN, JAMES D -XXXX35900 -L148/OS2SL -P3S3-92/05/12-20:53 -CR Re-queuing to bpcFr1. Priority P3. =DIMOCK, KENN L -XXXX35900 -L148/BPCFR1-P3S3-92/05/13-12:05 -CR safa at bcrvmpc1 is working this PMR -IBMSERV1 -XXXX35900 -L165/-------P3S3-92/05/13-15:06 -AT CALL DISPOSITION: __CLOSE X_FOLLOW UP NEEDED __NO CONTACT TIME SPENT WITH CUSTOMER: 00:30 SYSTEM CONFIGURATION: ZEOS 486, WITH DIAMOND SUPER VGA WITH ET4000 OS/2 1.3 DRIVER . QUESTION/PROBLEM: 1. HOLDS DOWN ALT & DRAGS 'NEW SCHEME PALETTE' TO DESKTOP. EXPECT SYSTEM WIDE CHANGE. FINDS THAT MENU HIGHLIGHT COLOR NOT SET. 2. FILE SYSTEM CRITERIA IN FOLDER SETTINGS-INCLUDE PAGE IS CHANGED. ON SHUTDOWN & REBOOT CHANGED INFORMATION IS LOST (FAT OR HPFS). CONCLUSION: NEEDS FOLLOWUP, FREQUENTLY WORKING FROM HOME AT (205)-672-XXXX. . KEYWORDS: PALETTE, SETTINGS --- The above information was received from SAFA AT BCRVMPC1 --- -IBMSERV1 -XXXX35900 -L165/-------P3S3-92/05/13-15:07 -AL -JOHNSTON, MICHAEL S. -XXXX35900 -L148/BPC8 -P3S3-92/05/13-15:10 -CR -CORWIN, PEGGY -XXXX35900 -L148/BPCFR1-P3S3-92/05/14-11:29 SCE ATTN SAFA --- DAVID CALLED TO OFFER INFO-- OS220 DRIVER FOR VIDEO CARD--AND PROBLEM STILL EXISTS. -SCHELL, AUSTIN -XXXX35900 -L148/BPCFR1-P3S3-92/05/20-18:34 SCR Requed for further asst. -CURRAN, KIM CE Y55-XXXX35900 -L148/OS2SLI-P3S3-92/05/20-20:08 SCE PLS CALL DAVID WITH STATUS. THANKS -RAYBURN, LINDA D. -XXXX35900 -L148/-------P3S3-92/05/21-09:19 -AL re-route to wpshl since it belongs there and the secondary was already routed there. -RAYBURN, LINDA D. -XXXX35900 -L148/WPSHL5-P3S3-92/05/21-09:22 SCC closing secondary (which got to WPSHL queues before primary). -RAYBURN, LINDA D. -XXXX35900 -L148/WPSHL2-P3S3-92/05/21-09:23 SCC closing secondary. -RAYBURN, LINDA D. -XXXX35900 -L148/INSTL2-P3S3-92/05/21-09:24 -CR re q to WPSHL where it belongs. -ARTZER, TRACEY -XXXX35900 -L148/OS2SL5-P1S3-92/05/26-09:35 SCE david would like a lev 2 cb on this sev 1 problem -LOVBLOM, GUNNAR S. .-XXXX35900 -L148/OS2SL5-P1S3-92/05/26-12:01 SCT Uppdate: David has this problem with his desktop: 1. Create a folder on the desktop,named TEST 2. Make a shadow of f.ex. the DOS folder 3. Include the .exe files only 4. Now the test folder contains all the *.exe only. After a shutdown and a reboot the test folder is a shadow of the whole Dos folder, not only the *.exe files. is there a bug in the shadow pgm ? ReQ to WPSHL on request of mr. WILKINS for imm. help. -RAYBURN, LINDA D. -XXXX35900 -L148/WPSHL2-P3S3-92/05/26-15:43 -CR re q to wpshl -RAYBURN, LINDA D. -XXXX35900 -L148/-------P3S3-92/05/26-15:43 -AL closing secondary -GRACHECK, BARBARA Y63-XXXX35900 -L148/OS2SLB-P2S3-92/05/28-08:01 SCE call cust. asap. linda rayburn has been working on problem. ph gd. -ANDERSON, FRANK -XXXX35900 -L148/OS2SLB-P2S3-92/05/28-15:02 SCR busy -IBMACE -XXXX35900 -L165/-------P3S3-92/05/29-12:00 -AL -HICKS, SONJA W53-XXXX35900 -L148/OS2SLE-P2S3-92/06/01-08:04 SCE david req os220 support =GLEASON, TERESA K -XXXX35900 -L148/OS2SLE-P2S3-92/06/01-17:14 SCC CLOSING SECONDARY. =GLEASON, TERESA K -XXXX35900 -L148/OS2SL4-P2S3-92/06/01-17:16 SCR REQUEUING SECONDARY TO LOCATION OF PRIMARY WPSHL. -REICH, DAVE .-XXXX35900 -L148/WPSHL -P2S3-92/06/02-08:32 SCC -RAYBURN, LINDA D. -XXXX35900 -L148/WPSHL -P2S2-92/06/03-16:08 -CR Raise sev, customer called recently to update. -RAYBURN, LINDA D. -XXXX35900 -L148/-------P2S2-92/06/10-11:02 -AL David described his problem to me as follows (PLEASE DON'T ASK him to rehash this, as he has already been asked multiple times): 1) Create a folder on the desktop. 2) Drag from one of the drives a directory containing .exe files, and other files, to this folder as a shadow. 3) From the Include Page of the settings notebook change the shadow to display only .exe files. (this should work). 4) Shutdown and reboot (I assume he shutdown properly, but did not specifically ask him that,darn it!) He then found that the include settings were no longer in affect for his shadow, he was now getting ALL files displayed, not just .exe files. I did not get these same results, my include filter remained in affect displaying only .exe files. David needs a callback on Monday 6/15 telling hime we could not reproduce it and either closing call or getting more specific info from him. He did mention, by the way, that he was able to recreate this on FAT and HPFS. -RAYBURN, LINDA D. -XXXX35900 -L148/-------P2S2-92/06/16-15:20 -AT Not in today, try tomorrow. -RAYBURN, LINDA D. -XXXX35900 -L148/-------P2S2-92/06/16-15:21 -AL -DITZIG, LINDA D. -XXXX35900 -L148/-------P2S2-92/06/22-14:30 -AT Started to tell David I could not reproduce his problem, but he got called away on another call. Would like someone to call him back today or tomorrow. -DITZIG, LINDA D. -XXXX35900 -L148/-------P2S2-92/06/22-14:31 -AL 1) Create a folder on the desktop. 2) Drag from one of the drives a directory containing .exe files, and other files, to this folder as a shadow (CTRL-SHIFT-BUTTON2). 3) From the Include Page 3 of the settings notebook change the shadow folder's settings by "ADDing criteria" by pressing the ADD pushbutton and selecting: Property to be compared - Real Name Comparison Type - Not Equal Comparison File - *.EXE Use of Criteria - AND Press the ADD pushbutton again 4) Close the settings notebook and the shadow folder will no longer display .exe files. This should work. 5) Now shutdown and reboot, look at your shadow folder again and the .exe files are back! This is easily recreated on HPFS and FAT. Creating APAR. -DITZIG, LINDA D. -XXXX07701 -L148/-------P2S2-92/06/22-15:44 -AL FUP for APAR resolution PROBLEM WAS INACTIVE FOR MORE THAN 14 DAYS -- CALL GENERATED -DITZIG, LINDA D. -XXXX07701 -L148/WPSHL -P2S2-92/07/08-18:01 -CC FUP for APAR Screen team 7/15/92; approved for fix Bumping severity up to 2 because of dup defect 55403, which was already a 2. Function reported as missing since r202; Borg ship issue Changed originator from M. Lucido to L. Blalock. As per Roberta Carey, rejecting defect as not correcting problem. CMVC:Fixed incorrect AFLD2 field. ═══ 2.3. *55823 Invalid association for *.INF prevents use of .INF files in WP ═══ name 55823 reference abstract Invalid association for *.INF prevents use of .INF files in WP duplicate state closed severity 3 answer perd-Prob APAR symptom vis duplicate defects: duplicate features: none. notes: The system ships with an invalid association for *.INF files that attempts to open the Command Reference with the new .INF file passed as a topic. Naturally, the topic is reported as not found. Users have tried to make a usable association, however, the predefined one always supercedes whatever they define. And it is not obvious where to go to delete the existing association. You have to open the notebook for the Command Reference in the Information folder in order to delete the erroneous association. The system should ship with a correct association or with no associate for *.INF files. Peter Magid is also aware of this problem. Screen team Approved File comm20.rc line 13, removed association of *.inf with view.exe if using param cmdref.inf. Also added new line to associate view.exe with *.inf. Charlene is going to delta this into r205 and r206. Instructed me to delta r202. Awaiting permission to do so. ═══ 2.4. *55880 SVT REGR:Unable to search with * and \* characters ═══ name 55880 reference abstract SVT REGR:Unable to search with * and \* characters duplicate state closed severity 3 answer pwa-Work Around symptom vis duplicate defects: duplicate features: none. notes: Driver: 6.463 Hardware: any Testcase: REFTEST There is a problem with the search function in the Command Reference and the Start Here applications. Start Start Here and select the SEARCH option. Type "OS\2\*" and press Enter. A message comes up that says the search text can't be found but the text panel that starts with the program has OS\2* in it. You can search this section or all sections and it will never find an asterisk in Start Here. The help text for search says that to find an asterisk use \*. However, this string search works in the books(CMDREF). The second problem with search only occurs in the books. Start the Command Reference and search for "win*" in hopes of finding all pages that have a word starting with "win"(windows, etc). The asterisk is supposed to be a wildcard but when it is used, the search is unable to find the text. This string combination works in the Start Here application. I attempted to recreate the above two problems on driver 6.472 with the following results: The first problem was not reproducible. Searching for both \* and os/2\* in all sections of Start Here turned up all occurrences of the strings (About OS/2 2.0 and Help Library: Description panels). The second problem was reproducible. For example, search through all sections of Command Reference (or REXX Information) for parameters and note the many occurrences. Then search through all sections for parameter and note the many occurrences. Now search through all sections for paramet* and note that no occurrences are found. I tested starthere and command reference on 6.474 and was not able to get any of the \* to work. In both books, I received search text not found. Lisa Madera ******************************************************* Sorry, I realized later that I had a typo. I could not get the * alone to work with win as in win*, but I was able to get it to work with os/2\* and win-os/2* when I had the correct slash in it. Lisa *************** This problem was caused by using DosCaseMap to build the CLVT but using WinUpperChar as the basis for comparison against that CLVT table. They appear to return different results. Therefore the fix is to use DosCaseMap for both. Since this problem occurs on r202, r205, and R206 - I am putting the fix in all three places. Oh, I forgot to mention - this is only a fix for the second problem. The first one, as noted by both Dave Blaschke and Lisa Madera is not reproducible if one uses the correct '/'. he search works correctly. The reason the second one failed for books, but not Start Here is because the characters which caused the table to be sorted out of order were not in Start Here. It IS a regression from 1.3, but not a regression since GA. I'm not sure what caused the regression since the code involved has not been touched since 1989. I suspect that WinUpperChar has changed its results, but I'm not going to try to prove it. Per request from Russ Blaisdell in Technical Support, I am attaching APAR PJ07120 to this defect since the customer reported the same problem as reported in this defect. This will enable that customer to get the fix. ═══ 2.5. 57818 *** ServicePak Regression *** ICON Creeps Up Desktop ═══ name 57818 reference abstract *** ServicePak Regression *** ICON Creeps Up Desktop duplicate state verify severity 2 answer chkm symptom vis duplicate defects: duplicate features: none. notes: I have Win 3.0 Solitaire set up to run as a seamless Windows application on my XGA (high res.) display. If I run it, then repeatedly minimize to an icon, restore, minimize etc, the position of the minimized icon gradually rises on the screen each time I minimize. This is only a few pixels at a time, but if you do it enough, it rises significantly. The horizontal position seems fine. THIS PROBLEM WAS FIXED IN THE GA RELEASE, BUT HAS REGRESSED IN THE SERVICE PAK INSTALLATION. THE COMPONENT SHOULD BE PMVIOP. SEEN ONLY ON XGA AND 8514/A VIDEO SYSTEMS. LOCAL FIX: Minimize to Minimized Window Viewer instead of the desktop. Screen team Approved Delegated to me by Dan Gisolfi Currently being worked on by Bill Sterner. Delegated. Change Team: Delegated work item! Delegated Icon is creeping up the desktop. Showed this problem to Ivan and he said that this was probably a shield problem and to reroute it to his team. A fix for this defect is currently being implemented. it requires changes to the PM Shield, WINOS2 Shield, and User (WINOS2 Window Manager). This defect has been fixed for winos230_r205 and r205. The fix for this defect required changes to the PM Shield, the WINOS2 Shield, and User (pmviop.dll, pmvdmp.dll, winsheld.exe, winsmsg.dll, user.exe). All three of these components must be built and used in sync!!! If there are any questions or if a private is required please contact Bob Zybach, OS/2 Level 3 Change Team, 3-XXXX. P.S. Mark Rogalski will be implementing these changes into the win31 and r206 releases. ═══ 2.6. 57928 Notebook BKM_TABTEXTLEFT Does not work ═══ name 57928 reference abstract Notebook BKM_TABTEXTLEFT Does not work duplicate state verify severity 2 answer asni symptom vis duplicate defects: duplicate features: none. notes: Within the pmstddlg.h BKS_TABTEXTLEFT is set to 0x0000, then within DrawTabText() (Within bookdtab.c line 963) appears the following: if (pBookInstance->ulNotebookStyles & BKS_TABTEXTRIGHT) fsCmd = (DT_RIGHT | DT_VCENTER | DT_MNEMONIC); else if (pBookInstance->ulNotebookStyles & BKS_TABTEXTLEFT) fsCmd = (DT_LEFT | DT_VCENTER | DT_MNEMONIC); else fsCmd = (DT_CENTER | DT_VCENTER | DT_MNEMONIC); as you can see the test for TABTEXTLEFT will never pass. Screen team Approved I am currently working on this defect BKS_TABTEXTLEFT constant incorrectly set to zero in header ═══ 2.7. *58193 SVT:REGR:Wrong context menu for Printer Icon in Driveduplicate ═══ name 58193 reference svt abstract SVT:REGR:Wrong context menu for Printer Icon in Driveduplicate state closed severity 3 answer perf-Prob FC symptom vis duplicate defects: duplicate features: none. notes: OS/2 Level : 6.475 From Drives Folder, select Drive A and insert a printer driver disk in Drive A. All the Printer drivers on the disk are displayed now. Double-click on one of the printer driver icons. Now all the printer models associated with the driver should be displayed. But instead, the OS/2 System Editor is displayed. The context menu for the Printer Icon should have Open - Settings, Icon View. Instead, it has Open - Settings, OS/2 System Editor, with the Editor selected, and that's why we do not get the printer models. Talked to Kim Shepard and Cathy Bloomfield about this problem. Customer Impact : Accessing the printer models from the drives Folder and dragging and dropping the printer neeed by the user onto the Desktop is one of the ways of creating Print Objects. Due to the problem mentioned above, the Print Object cannot be created through this method. Vish - Printer Lab 2-XXXX I looked at this briefly... our parent class is WPDataFile. We used to override wpclsQueryIcon, wpclsQueryStyle, wpMenuItemSelected, wpFilterPopupMenu wpModifyPopupMenu, etc. We would return back the right menu, icon, etc. when you opened the drive and directory with printer drivers in it. Now, I don't hit those breakpoints at all... This worked in 6.474 and was broken in 6.475. I still have a 6.475 build report if you want to see what you delta'd. We didn't change anything. Honest! I still have 6.474 on my test machine if you want to look at it. I'm happy to look at it with you... just come and get me. Thanks, Kim S. ═══ 2.8. *58329 Most 50MHz Phoenix BIOS do not boot. ═══ name 58329 reference none abstract Most 50MHz Phoenix BIOS do not boot. duplicate state verify severity 3 answer dsam symptom vis duplicate defects: duplicate features: none. notes: Most 50MHz Phoenix BIOS do not boot. This problem has also been found on Ambra machines with the latest Phoenix BIOS. These machines are not 50 MHz. Paul. Screen Team Approved Fixed by Russ. Real time clock save and restore needed on real/protect switch. Never hit before as we were on d1 before entering prot mode, noe we enter prot mode on in disk. Paul. We had 3 machines with this problem in the OEM Lab on driver 6.494. Pamela Marshall 982-XXXX (v$ipamm) OEM Lab It is totally unlear to me as to why the OEM Lab added the above comment - this fix is in 496, so it is not surprising that it fails on 494. Paul. We have this problem on 2 machines on driver 6.498a in the OEM lab. Pamela Marshall (v$ipamm) 982-XXXX Correction, We did not have this problem in the OEM lab. Pamela Marshall (v$ipamm) 982-XXXX ═══ 2.9. *59295 SVT: Query manager extremely slow to open ═══ name 59295 reference abstract SVT: Query manager extremely slow to open duplicate state verify severity 2 answer lfi symptom other duplicate defects: duplicate features: none. notes: A loader problem is creating a consistent loop for the Query Manager. Therefore, the Query Manager is extremely slow. It takes a few minutes before access to the Query. A second attempt to open the Query Manager sometimes causes the system to hang. This seems to be a loader problem and Paul Devriendt is currently working on this defect. Please call Ann-Charlotte Gruber at 2-XXXX if additional information is required. Fixed in 6.486 build with performance enhancements to kernel (ldr/tk/vm) and to os2char.dll/pmwin.dll/doscall1.dll/sesmgr.dll. Paul. ═══ 2.10. *59304 This is a regression from GA - Notebook tabs unreadable ═══ name 59304 reference PJ06844 abstract This is a regression from GA - Notebook tabs unreadable duplicate state closed severity 4 answer dsai symptom vis duplicate defects: duplicate features: none. notes: When attempting to change font size from Desktop settings, the notebook tab text is increased in size accordingly. This makes the names on the tabs unreadable if the characters are too large for the tab. This does not occur in GA. To recreate the problem do the following: 1. Click right mouse button on desktop and open settings notebook. 2. Click on right arrow twice to get to page 3 of 3. 3. Click on Change Font. 4. Change font size to Times 24. 5. You will notice the notebook tab text has changed in size accordingly and will be unreadable. For more information contact Ajit Abhyankar at (2-XXXX) or John O'Neil (2-XXXX). Screen team Approved This problem is in the workplace handling of the desktop notebook pages. The notebook page is assigning the font to the notebook tabs instead of to the Details View display! The proper usage of the font information needs to be done in the workplace code, not in the graphics engine or fonts. While you are looking at this, it also seems that if you are changing the font for the tree view, it changes the font for the icon view as well. James ═══ 2.11. 59344 When system is locked and alarm goes off; alarm on top of locku ═══ name 59344 reference abstract When system is locked and alarm goes off; alarm on top of locku duplicate state working severity 3 answer inti symptom use duplicate defects: duplicate features: none. notes: To recreate the problem I followed the customer's steps one by one: 1.- Double click on the alarms icon 2.- Left click on ALARMS-SET ALARM 3.- Number = 0 4.- Status = ON 5.- Day = Every Day 6.- Hour = Every Hour 7.- Minute = 00 8.- Graphic = None 9.- Comment = Leave it blank 10. Tune = 0:Default Tune 11. Activate Every Alarm time occurrence 12. Select PopUp ** ( YOU MUST SELECT POPUP FOR THE PROBLEM TO OCCUR 13. Left click on SET 14. From the OS/2 window set the time to be 12:59 15. Lock the workstation 16. Wait for the alarm to go off 17. Type an incorrect password ( The alarm popup shows on top of the lockup screen ). Screen team Approved Change Team: Delegated work item! Adding the APAR field which was missing. Modified. This defect is available; CT played ping-pong before APAR info existed Changed the lockup to replace the original Message Box that warned of an incorrect password being typed in, instead coded a Dialog Box. The thought was that a dialog box can 'grab' the focus and a message box can't. However, the Alarm Window that comes in on top of the incorrect password notification is sysmodal. The dialog box that I coded is also sysmodal. AND the system doesn't handle multiple sysmodal windows very well. SO...there could be a major design change here. Am awaiting James Taylors return from a conference next week. ═══ 2.12. 59480 BETA:WIN_RUNMODE description is incorrect ═══ name 59480 reference abstract BETA:WIN_RUNMODE description is incorrect duplicate state canceled severity 3 answer invdup symptom use duplicate defects: none. duplicate features: none. notes: BETA:Bring up the WIN-OS2 settings in the settings notebook for a windows app. The WIN_RUNMODE parameter has changed from the GA version but the description field has not been updated to reflect the changes. This is a duplicate of defect 59574 already written by Mel. -- Ayo Anise (2-6794) ═══ 2.13. *59509 BETA:SYS3006 on AUTOEXEC.BAT DOS settings change ═══ name 59509 reference abstract BETA:SYS3006 on AUTOEXEC.BAT DOS settings change duplicate state closed severity 3 answer dsai symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA:If you change the AUTOEXEC DOS setting in a DOS Full Screen session to point to your batch file, you get SYS3006 when the session starts. The SHELL statement in the CONFIG.SYS and the DOS_SHELL statement in the DOS settings correctly point to the path for the command.com. User advises this function did work in GRE-WIN earlier driver. (John Beidl in Raleigh) I've added parameter passing to the DOS_AUTOEXEC. The routine that stepped through the parameters had a bug. It should now work. ═══ 2.14. 59512 BETA: User has pointed out that the tutorial still refers tduplicate ═══ name 59512 reference abstract BETA: User has pointed out that the tutorial still refers tduplicate state canceled severity 3 answer valunr symptom vis duplicate defects: none. duplicate features: none. notes: BETA: User has pointed out that the tutorial still refers to OS/2 2.0 instead of OS/2 2.1 All tutorial text changes are made through ID. Roberta Carey 3-XXXX We have changed the tutorial to reflect the OS/2 2.1 be in the next beta driver. This defect was corrected under another defect number. ═══ 2.15. *59514 BETA:shutdown dialog box refers to 2.0 instead of 2.1 ═══ name 59514 reference abstract BETA:shutdown dialog box refers to 2.0 instead of 2.1 duplicate state closed severity 3 answer inti symptom vis duplicate defects: none. duplicate features: none. notes: BETA: User has reported that the SHUTDOWN dialogue box refers to 2.0 instead of 2.1 Please transfer this to the correct component if this is the wrong one. I do not understand this defect. THere is no reference to os/2 versions in the shutdown dialog box or any other part of shutdown. please explain further. --Mindy Pollack 12/23/92-The 2.0 I was referring is at the top of the Shutdown Dialogue when you request to shutdown. Users reported they thought it should be 2.1 VAL, The problem here is the title of the desktop (in the 20 rc file) is called OS/2 2.0 Desktop. We agreed with usability to change it to "Desktop". (leave OS/2 1.3 Desktop and WIN 3.1 Desktop the way they are). Please send a note to me when you have completed this defect so I can tell ID that it is definite. --Mindy Per Dylliam approval this MRI change is acceptable. The title of the desktop has been changed from OS/2 2.0 Desktop to Desktop. ═══ 2.16. *59530 BETA:Running HELP in PAX/PM folder disables help ═══ name 59530 reference abstract BETA:Running HELP in PAX/PM folder disables help duplicate state closed severity 3 answer inti symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: User reported and I verified that with the FAX/PM folder open, if you RMB on an icon and select help and then general, it works okay. But if you now do the same thing to another icon in that folder, help seems to get disabled everywhere. You click on it and nothing happens. I can only get the error when I use the FAX/PM folder. In other folders, the first help screen keeps getting overlayed by the latest. This folder seems to be working differently. Please transfer this to the correct component if I made the wrong choice. Bob Rosenman 3-XXXX Please verify resid's with Anita Soto. The Fax/PM viewer and Fax/PM Scheduler icons should be linked to Master help index. I was able to recreate the problem by hitting F1 on Fax/PM viewer icon and afterwards it seemed all the helps were disabled. 1/8/93 (Val Robinson) The correct resid's were used for the Fax/PM viewer and scheduler(21500) and the correct helplibrary name (WPINDEX.HLP) was also used to enable the help screens to work properly. ═══ 2.17. 59536 BETA:SYS3175 on TYPE with MORE misspelled ═══ name 59536 reference abstract BETA:SYS3175 on TYPE with MORE misspelled duplicate state canceled severity 3 answer invdup symptom fnwad duplicate defects: duplicate features: none. notes: BETA: User has reported and I can verify the following error: Enter- type a:readme.doc | mroe (more is misppelled) I get the correct error message that it doesn't know "mroe" command. Reenter the command with "more" spelled correctly. You get SYS3175, SYS1808 with exception 0005. Run from OS/2 window. Bob Rosenman 3-XXXX Please transfer to correct component if this is the wrong one. Assigning to proper component. Duplicate of 59591 in mkio component. ═══ 2.18. *59550 BETA: DIR /S figures the bytes used incorrectly ═══ name 59550 reference abstract BETA: DIR /S figures the bytes used incorrectly duplicate state closed severity 3 answer asni symptom fnwad duplicate defects: duplicate features: none. notes: BETA:Users have reported and I have verified that the DIR /S command is showing a very high incorrect value for the bytes used field both in HPFS and FAT partition and independent of the size of the partition. Please transfer this to the correct component if I have indicated the wrong one. Assigning to proper component. Correct errors in bytes used logic and print out free bytes for each drive when a change of drives occur from multiple dir parameters. ═══ 2.19. 59552 BETA: Use of HELP in FAX/PM folder disables HELP in system ═══ name 59552 reference abstract BETA: Use of HELP in FAX/PM folder disables HELP in system duplicate state canceled severity 3 answer symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA:User reports and I have verified that if you open the FAX/PM folder in the the productivity folder and do a RMB on any of the 3 icons and select help and then general help, it works okay. If you then do the same thing on another of the icons in that folder, the help function seeems to be disabled from that point on. You can select help but see nothing. In similar operations in other folders, the latest requested help overlays the current one, but it doesn't seem to work in this folder. I thought I had submitted this problem yesterday as number 59530 but now I can't find it in the system, so I am resubmitting. Please transfer to correct component if I've indicated the wrong one. Bob Rosenman 3-XXXX This is a dupe of 59530. Problems in CMVC data base concerning the dev owner of this component caused me to think 59530 was lost. After the CMVC folks straightened things out, I saw this is a dupe so I am cancelling it. ═══ 2.20. 59555 BETA:Trap E copying OS/2 clipboard then pasting WINOS2 clipduplicate ═══ name 59555 reference abstract BETA:Trap E copying OS/2 clipboard then pasting WINOS2 clipduplicate state canceled severity 3 answer valunr symptom thipe duplicate defects: none. duplicate features: none. notes: BETA:User reported following TRAP E which I have reproduced- Bring up OS/2 window and do a COPY ALL from the system menu. Then start WINOS2 full screen and bring up the notebook applet. Do a paste. Trap E results. Easy to reproduce. Contact me if you need any specific Trap info. Please transfer to correct component if this one is incorrect. Bob Rosenman 3-XXXX This is not a WinOS2 issue but a viow issue. The dialog box in question is in viow. Darren & GlenP I no longer see the problem in this latest driver (6.481) ═══ 2.21. *59576 BETA:FAX/PM help missing for modem tuning in notebook ═══ name 59576 reference abstract BETA:FAX/PM help missing for modem tuning in notebook duplicate state closed severity 3 answer docm symptom use duplicate defects: none. duplicate features: none. notes: BETA: User has reported and I have confirmed that the help for the FAX/PM Schedular applet is missing help info for the Modem Tuning page in the notebook under Communication then Setup. The Setup (under Communications then Setup) option allows you to change certain parameters which are listed under Miscellaneous Options, and Receive Options. Modem Tuning is not an option that you can change and therefore is not listed on the Fax/PM Setup help panel. The Modem Tuning help panel discussing SEVERAL OPTIONS that you CAN change. To get help for modem tuning, follow the instructions below. - While in the Scheduler, select communication. - Select setup - Select any field (ie. Scan time) by pressing mouse button one. - Press F1 You will get the Modem Tuning Help Panel. Thank you, Anita Soto (2-XXXX) I've spoken with Anita and I believe we agree that the Modem Tuning help exists, but a link to it should be added to the FAX/PM Setup help screen. This will allow users to be able to see it when they click on the Help button that is on the Modem Tuning page or any of the previous tab pages in Setup. I have spoken with Anita and I believe that we agree that the Modem Tuning help exists but a link to it should be added to the Fax/PM Setup Help screen. This will allow users to see the help when they click on the Help button that is on the modem tuning tab or on any other tab in the Setup section. 59576 Working. Link to Modem Tuning is being added to the Fax/PM Setup page. Anita Soto (2-XXXX) 59576 Working. Link to Modem Tuning is being added to the Fax/PM Setup page. Anita Soto (2-XXXX) ═══ 2.22. 59578 BETA: SYS3175 access violation on MIGRATE of MS Bookshelf ═══ name 59578 reference abstract BETA: SYS3175 access violation on MIGRATE of MS Bookshelf duplicate state canceled severity 3 answer valunr symptom thipe duplicate defects: none. duplicate features: none. notes: BETA: User has reported SYS3175 access violation during migrate of MS Bookshelf. We also see the problem migrating MS Publisher While migrating MS bookshelf I get "SYS3175 a program in this session encountered a problem and cannot continue. Displaying the reg info it says"A program generated an access violation at 00010181 MIGRATE.EXE 0001:0000181. Viewer.exe is the file name that is being matched in the database.dat This is the first and only match in the database. To prove this I created a new database with just bookshelf in it (viewer.exe etc) and it still failed. I have successfully migrated OS/2 and DOS programs but none that are in the database. It looks like the problem only happens when there is a match in the database. Any ideas? (I'm USING the latest OS2/ 2.1) This defect will be unreproducible on next version. ═══ 2.23. 59579 BETA:STARTUP.GRP path wrong in PROGMAN.INI ═══ name 59579 reference abstract BETA:STARTUP.GRP path wrong in PROGMAN.INI duplicate state closed severity 3 answer bldi symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: User reports and we have verified that when OS/2 is installed on other than the C drive, the PROGMAN.INI file has the incorrect path for the STARTUP.GRP file. The path points to the C drive, although the user has installed OS/2 Beta on another drive such as D. Boot Manager was used to run the beta from the other drive. User was getting GROUP FILE ERROR. ═══ 2.24. 59582 BETA: Icons not recognized by find utility ═══ name 59582 reference abstract BETA: Icons not recognized by find utility duplicate state canceled severity 3 answer symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: A customer has sent me 104 icons that the find utility will not recognize as icons unless they have been loaded into the icon editor and then saved. GA + SP does not ═══ 2.25. 59583 BETA: Icons not recognized by find utility ═══ name 59583 reference abstract BETA: Icons not recognized by find utility duplicate state canceled severity 3 answer symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: A customer has sent me 104 icons that the find utility will not recognize as icons unless they have been loaded into the icon editor and then saved. GA + SP does ═══ 2.26. 59584 BETA: Icons not recognized by find utility ═══ name 59584 reference abstract BETA: Icons not recognized by find utility duplicate state canceled severity 3 answer symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: A customer has sent me 104 icons that the find utility will not recognize as icons unless they have been loaded into the icon editor and then saved. GA + SP does find them correctly without having to load them into the icon editor and save them first. The steps to reproduce this are: 1) load icons into subdirectory called OS2ICONS ═══ 2.27. 59585 BETA: Icons not recognized by find utility ═══ name 59585 reference abstract BETA: Icons not recognized by find utility duplicate state canceled severity 3 answer symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: A customer has sent me 104 icons that the find utility will not recognize as icons unless they have been loaded into the icon editor and then saved. GA + SP does find them correctly without having to load them into the icon editor and save them first. The steps to reproduce this are: 1) load icons into subdirectory called OS2ICONS ═══ 2.28. 59586 BETA: Icons not recognized by find utility ═══ name 59586 reference abstract BETA: Icons not recognized by find utility duplicate state canceled severity 3 answer symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: A customer has sent me 104 icons that the find utility will not recognize as icons unless they have been loaded into the icon editor and then saved. GA + SP does find them correctly without having to load them into the icon editor and save them first. The steps to reproduce this are: 1) load icons into subdirectory called OS2ICONS ═══ 2.29. 59590 BETA: Find utility does not recognize icons ═══ name 59590 reference abstract BETA: Find utility does not recognize icons duplicate defect 58193 state canceled severity 3 answer duplicate symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: A customer has sent 104 icons that the find utility will not recognize as icons unless they have been loaded into the icon editor and saved first. They are recognized without needing to be loaded and saved using GA+SP. Other icons that I have tried are found in GA+SP as well as 6.479c. The problem seems specific to these icons. The following are the steps to reproduce this problem 1) Create a new objuct from templates, both folder and program object have been tried. 2) Go to general page of new object 3) select FIND 4) select LOCATE 5) select PATH and enter path to icons 6) select OK 7) select FIND No objects are found. If one opens the Icon editor, loads one of the icons, and saves it without changeing anything, then the above steps will find the icon that was saved and still refuses to recognize any that have not been saved by the editor. Please contact me for the icons that are causing the problem or if you have any questions at phone# 2-XXXX. The cause of this problem is the same as what caused 58193, which has been fixed and will go into the next post-beta build. We have tested this scenario with a private fix for 58193 and it works. Sheila 2-XXXX ═══ 2.30. *59591 BETA:SYS0447 and SYS1805 on DIR /P, ctl-break,type xx.xx ! duplicate ═══ name 59591 reference abstract BETA:SYS0447 and SYS1805 on DIR /P, ctl-break,type xx.xx ! duplicate state closed severity 3 answer timm symptom fnwad duplicate defects: duplicate features: none. notes: BETA: Problem reported by user and reproduced by myself-Use os/2 window Do a DIR /P for a directory that is more than one screen full of data. When it pauses, do a ctrl-break. Now do a type config.sys | more - Right after the more shows up at bottom of screen, you get SYS0447-The keyboard focus is not available SYS1805-The process tried to write to a nonexistent pipe. Please transfer this to correct component if I have indicated the wrong one. There is an easier way to reproduce this problem. Simply hit CTRL-BREAK at the command prompt (that's right, at the [C:\] prompt), then try "MORE < CONFIG.SYS". The SYS0447 error will occur after one pageful of data, and the MORE command will not work at all for the remainder of the session. This problem occurs in both the OS/2 FullScreen and OS/2 Window. -- Kevin J. North, LAN Lab 3-2805, KNORTH@BCRVMPC1 Possible duplicate of 59536. The problem was that the per session semaphore was set as owned by KbdCharIn() and never cleared under the following conditions. - Run a full screen protect mode session - Hit - Use any program that makes keyboard API calls CMVC:Fixed incorrect AFLD2 field. CMVC: cloned to new defect 61793. ═══ 2.31. 59595 BETA: Set DELDIR Statement Causes System Hang On Deletes ═══ name 59595 reference abstract BETA: Set DELDIR Statement Causes System Hang On Deletes duplicate state canceled severity 3 answer invdup symptom thipe duplicate defects: none. duplicate features: none. notes: With Statement "SET DELDIR = x:\DELETE, 512; active using the drive Icon and then going to a directory and deleting multiple icons causes the system to halt requiring the user to power off the machine to recover. Remming out the statement in the config.sys file makes the problem go away. "REM SET DELDIR = X:\DELETE,512; You must debug the problem before you submit a defect. There is only a creation scenario in this problem. Please check your problem against the first build after the beta, it should be the build that will be created this week. There was a fix for undelete that was not put into the beta, so check your scenario against the proper build. G. Simco Reopening per my conversation of 1/06/93 with Greg. This problem is a duplicate of defect 57983, which was not built in the beta. I tested the submitters scenario on build 481 without a problem. G. Simco SInce this is a duplicate of a defect that wasn't fixed in the beta, and has been confirmed by Greg as working in the next build I am going to cancel this defect. ═══ 2.32. 59596 SET DELDIR Statement Causes System Hang on Deletes ═══ name 59596 reference abstract SET DELDIR Statement Causes System Hang on Deletes duplicate state canceled severity 3 answer symptom thipe duplicate defects: none. duplicate features: none. notes: With Statement "SET DELDIR = x:\DELETE, 512; active using the drive Icon and then going to a directory and deleting multiple icons causes the system to halt requiring the user to power off the machine to recover. Remming out the statement in the config.sys file makes the problem go away. "REM SET DELDIR = X:\DELETE,512; ═══ 2.33. *59597 BETA: DOS_INT28_FILTER missing from the DOS SETTINGS list ═══ name 59597 reference abstract BETA: DOS_INT28_FILTER missing from the DOS SETTINGS list duplicate state closed severity 3 answer doci symptom fnwad duplicate defects: none. duplicate features: none. notes: The online help for the "DOS_settings..." mentions that DOS_INT28_FILTER is a changeable setting. However, this setting is not shown in the list. The steps to get the paragraph where the DOS_INT28_FILTER setting is mentioned are: - Open the settings for any DOS session - Select the "Session" tab and press the DOS_settings... pushbutton - Select HELP - From the Related Information area, select "DOS AND WIN-OS/2 settings, descriptions" - Select "Other DOS Settings" I don't know what the INT28 filter setting is supposed to be. I can tell you it is not registered by any of the standard devices on our system as of today. You might ask Jeff Muir about this. I am re-routing this to BASEINFO because that component owns the file MRI\HELP\DOSPRMN, where the text in question is located. Accepted for r206 build on or about 1/18/92. ═══ 2.34. *59610 BETA:Install 5202-3 printer, asks for diskette -48 ═══ name 59610 reference abstract BETA:Install 5202-3 printer, asks for diskette -48 duplicate state verify severity 3 answer bldi symptom fnwad duplicate defects: duplicate features: none. notes: BETA: Problem reported by a user and reproduced here- Do an install on a formatted partition of everything and specify for a printer the 5202 QW III. The install goes ok until after it has read printer driver diskette 3. Then it asks for printer driver diskette -48 and will accept none of the printer driver diskettes. Only action for user is to boot to get out of the problem. ═══ 2.35. 59619 BETA:VER says 2.00 instead of 2.10 ═══ name 59619 reference abstract BETA:VER says 2.00 instead of 2.10 duplicate state canceled severity 3 answer invunr symptom fnwad duplicate defects: duplicate features: none. notes: BETA: User has reported and I have confirmed that when you do a VER in a DOS window or full screen, it says the vesion is 2.00 instead of 2.10. BETA: Another user is now also reporting this problem for the OS/2 window. Command.com's VER is based on the value in h\dosver.h. I use the DOS_VERSION equate. A.Soto has already fixed this in Borg to Ver 2.1 ═══ 2.36. 59623 BETA:Reverse Sender indentifier case problem ═══ name 59623 reference abstract BETA:Reverse Sender indentifier case problem duplicate state canceled severity 4 answer valunr symptom vis duplicate defects: none. duplicate features: none. notes: BETA:On the Modem Tuning Screen, the last item- 'Reverse Sender identifier' should be 'Reverse sender identifier' to be consistent with the other entries on that screen and the way it is written in the HELP for that entry. (Note lower case s is the difference) The fix for this defect has been delta'ed in defect 60369. Please verify your defect when the driver 6.490 comes up. Thanks. Screen Team Approved ═══ 2.37. *59630 BETA:typo in DOS and WIN-OS/2 Video Mode Considerations ═══ name 59630 reference abstract BETA:typo in DOS and WIN-OS/2 Video Mode Considerations duplicate state closed severity 4 answer doci symptom vis duplicate defects: none. duplicate features: none. notes: BETA: User has reported and I have confirmed there is a typo in the DOS and Win-OS/2 Video Mode Considerations help panel. Search on difficulty to see it. It says 'It is difficulty to provide virtual access ...'. It should be 'difficult' not 'difficulty'. Open System, search help all sections for difficulty to recreate. Defect fixed will integrate next build. ═══ 2.38. *59631 BETA:CPU Pulse overruns bu 20% on DX2-66 system ═══ name 59631 reference abstract BETA:CPU Pulse overruns bu 20% on DX2-66 system duplicate state closed severity 3 answer dsai symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: User has reported this problem in the OS2BETA forum- ----------------------------------------------------------------- Subject: Pulse meter overrun I have reported this bug a while ago with the previous beta, and is still in the current beta. CPU Pulse overruns by 20% on a DX2-66 system. With nothing running, it is showing 80% CPU drain. Perhaps someone should convert pulse to run a Dhryston benchmark instead of a simple loop. This way, we should never encounter overrun problem again even when OS/2 is installed on a Pentium! Albert Shan OS/2 Ambassador / Vancouver ISC Level 2.5 OS/2 Support ------------------------------------------------------------------ I will interface between this user and the developer Bob Rosenman 3-XXXX Problem was an overflow on a multiply on fast machines. Solved by reversing the order of the multiply and divide, if the multiply would cause overflow. Also converted to C Set/2 instead of MS c386. ═══ 2.39. 59632 BETA: 8514 install is prompted for DSPLYDRV2 diskette ═══ name 59632 reference abstract BETA: 8514 install is prompted for DSPLYDRV2 diskette duplicate state canceled severity 3 answer valunr symptom fnwad duplicate defects: none. duplicate features: none. notes: Customer installing from CDROM, selects 8514 driver, installation then asks him for Display Driver Diskette #2 from F: (his CDROM DRIVE). Sorry - I'm rerouting this without reproducing it, just so it gets to the right component. Could not re-create this on 488. Duane Wood Canceled Per duanew ═══ 2.40. 59634 BETA:Deleting folder containing many folders hangs system ═══ name 59634 reference abstract BETA:Deleting folder containing many folders hangs system duplicate state canceled severity 3 answer valunr symptom thipe duplicate defects: duplicate features: none. notes: BETA: Many users have reported and I have verified that when you take a folder such as FIND RESULTS to shredder and tell it to shred, the system hangs and the folder and its contents never go away. However, if you go to the OS/2 command prompt and delete the directory with the FIND RESULTS, everything goes away in a flash. To recreate, RMB on desk top-click on find-click on search all folders- click on find and wait until results are displayed-close results window- drag find results folder to shredder. we need to do something like not delete folders with dirty children or have any of the children that are dirty refuse to delete because they are dirty. It might confuse the user but at least they won't trap --Mindy Screen Team Approved ═══ 2.41. 59635 BETA:OS/2 Window assigned icon not shown when minimized ═══ name 59635 reference abstract BETA:OS/2 Window assigned icon not shown when minimized duplicate state canceled severity 3 answer invsug symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: User has reported and I verified the following- -------------------------------------------------------------------- Subject: Custom icon not shown when window is minimized - assign an icon to say an OS/2 Window object - open the session and then minimize it - the minimized icon shows generic OS/2 windows icon I remember seeing a number of APARs talking about this problem, and all of them were closed, saying it'll be fixed in the next release. So far, it is not fixed yet. ----------------------------------------------------------- I don't know about ever saying this would be fixed in the next fixed in the next release we say it will be CONSIDERED in the next release. The design of the workplaces shell states that workplace just fires off other applications. Although you can set the icon of the OBJECT that represents the OS/2 window, the shell has no control over what the minimized window will look like because that is up to the application. Workplace has no interface for changing minimized windows. An OS/2 window is just another application and we have no control over what icon they use when the window is opened. In the future we are going to try to make the OS/2 windows work more closely with workplace to include all this but that will not be done for 2.1. --Mindy Pollack 2-XXXX I accept the response. ═══ 2.42. 59638 BETA: inappropriate INF association ═══ name 59638 reference abstract BETA: inappropriate INF association duplicate defect 55823 state canceled severity 3 answer duplicate symptom use duplicate defects: none. duplicate features: none. notes: BETA: Here's one that has propagated all the way from GA 2.0 After install, if one attempts to double click on a .inf file, the message Topic Not found is displayed. The problem is that .inf's are inappropriately associated with the command reference, which happens to be a specific instance of view.exe , rather than view.exe during install. Of course when one unassociates *.inf with the command reference and adds *.inf to the association page for view.exe, clicking on a .inf will open view.exe on that .inf. This is really a SHAPI problem but was fixed by defect 55823. duplicate of 55823 ═══ 2.43. *59648 Dos programs use of DMA causes unpredictable result. ═══ name 59648 reference abstract Dos programs use of DMA causes unpredictable result. duplicate state closed severity 2 answer dsai symptom thipe duplicate defects: duplicate features: none. notes: Customer is a developer trying to use DMA. It seems that when he issues the following command to the DMA controller: eax=00007201 ebx=00000000 ecx=00007192 edx=000000d4 esi=00000000 edi=00000136 eip=0000056f esp=00001f50 ebp=00001f66 iopl=3 -- vm -- nv up ei pl nz ac po cy cs=5894 ss=73a9 ds=72b8 es=72bf fs=0000 gs=0000 cr2=7c3b1110 cr3=001c5000 5894:0000056f ee out dx,al The system goes out of control. For example, in the following debug exerpt, a breakpoint is set immediately after the above OUT instruction. When the instruction is executed, the systems stops in the condition shown: 5894:0000056f ee out dx,al 5894:00000570 33c0 xor ax,ax eax=00007201 ebx=00000000 ecx=00007192 edx=000000d4 esi=00000000 edi=00000136 eip=0000056f esp=00001f50 ebp=00001f66 iopl=3 -- vm -- nv up ei pl nz ac po cy cs=5894 ss=73a9 ds=72b8 es=72bf fs=0000 gs=0000 cr2=7c3b1110 cr3=001c5000 5894:0000056f ee out dx,al --bp cs:570 --g eax=00007201 ebx=00000000 ecx=00007192 edx=000000d4 esi=00000000 edi=00000136 eip=00000570 esp=00001f50 ebp=00001f66 iopl=3 -- vm -- nv up ei pl nz ac po cy cs=5894 ss=73a9 ds=72b8 es=72bf fs=0000 gs=0000 cr2=7c3b1110 cr3=001c5000 Invalid address type: aa - 5555|5555:aaaa5555 I have a debug log of the above events, which contains the instructions used to set up the DMA transfer etc. Customer can reproduce any time. However due to special hardware requirements, we cannot reproduce in our lab. Please call Dennis Sposato 2-XXXX for more info. Reroute to VDMA. Screen team Approved VDMA has a problem with Channels 5, 6, and 7 on ISA machines. The code was not wrong by much, but enough to cause major trouble. Fixed it and the customer says his test worked. ═══ 2.44. *59654 BETA: BREAK not set as indicated in software configuration ═══ name 59654 reference abstract BETA: BREAK not set as indicated in software configuration duplicate state closed severity 3 answer inti symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: Users have reported that when they set BREAK either ON or OFF in Software Configuration/Change DOS Parameters during install, the config.sys that gets created always has the opposite BREAK value then what they set. The pop-up indicates ON is default, but they get OFF. If they specify OFF, they get ON. ═══ 2.45. 59673 Beta: ATM icon for winos2 incorrect ═══ name 59673 reference abstract Beta: ATM icon for winos2 incorrect duplicate state closed severity 3 answer bldi symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: after installation one notices the atm control panel icon in group main is incorrect. If one is to open the atm control panel all works well. However , if one highlights the icon selects File and then properties and attempts to change the icon one gets the error message that " the path c:\os2\mdos\winos2\atmcntrl.exe is invalid" OS/2 is installed on the logical F partition not the C. When one changes the path to the proper installation drive, the Icon becomes the large script a. This is an OS/2 install problem. Install should update the atm.ini file with the appropriate drive and directory information when OS/2 is installed on another partition other than C: -- Ayo Anise (2-6794) Actually the problem was in WOS2MAIN.GRP, which had hard-coded pathnames for ATMCNTRL.EXE and NOTEPAD.EXE. Removing the full pathnames corrected this problem. Don Wakley 2-XXXX ═══ 2.46. 59703 BETA: GDI.EXE UNABLE TO LOAD - CAN'T RUN WIN/OS2 FULL SCREEduplicate ═══ name 59703 reference abstract BETA: GDI.EXE UNABLE TO LOAD - CAN'T RUN WIN/OS2 FULL SCREEduplicate state canceled severity 1 answer invdld symptom appf duplicate defects: none. duplicate features: none. notes: DURING THE INSTALL PROCESS THE USER SELECTED AN OPTION TO PRESERVE THE WIN/OS2 DESKTOP. THE INSTALLATION WAS SUCCESSFUL BUT THEY COULDN'T RUN WIN/OS2 FULL SCREEN OR LAUNCH SEAMLESS APPS. AT ONE POINT AN ERROR MSG APPEARED - *GDI* WAS UNABLE TO LOAD. THERE HAVE BEEN MULTIPLE APPENDS TO THE OS2BETA FORUMS, BOTH BY INTERNALS AND ISV'S, RELATING TO THIS PROBLEM. FOR MORE INFORMATION CONTACT MARC SCHNEIDERMAN - BTIG 3-XXXX ROOM W-2035 The beta version of install was not adding the statment "os2gdi.exe=GDI.EXE" to the [boot] section of setup.inf. This has been fixed since the beta was released. I tested the install program built for the 6.482 build and it does add this statement to setup.inf. (Pardon my use of the "downlevel driver" code - I did not intend to imply that the problem was invalid, only that it has been fixed in a later driver). Correction: system.ini is the file that gets this statement, not setup.inf. Thanks to Bob Dutton for his sharp eyes. Don Wakley 2-XXXX. ═══ 2.47. *59704 BETA:PCAD autoload hangs the system, math coprocessor missiduplicate ═══ name 59704 reference abstract BETA:PCAD autoload hangs the system, math coprocessor missiduplicate state closed severity 3 answer chkm symptom thipe duplicate defects: none. duplicate features: none. notes: BETA: User has reported this problem and I have reproduced it here. The user has PCAD application which comes from CADAM, an IBM subsidiary. This application requires a math coprocessor. In the GA, if you try to install it on a machine without a coprocessor, you get SYS2237 which says A NPX instruction was attempted but no NPX is present. In the 6.479C driver, the system hangs and needs to be powered off to recover. Please contact either myself 3-XXXX or Mike Turner 2-XXXX to recreate. Please pass this on to another component if I have indicated the wrong one. Assigning to proper component. Screen Team Approved The 386 chip hangs if a floating point instruction is executed and a co-processor is not installed. The problem is that DPMI is allowing apps to physically turn off the EM bit through DPMI calls. The fix is to alter shadow bits instead of the physical bits. I gave the fix to Bill Madden to delta. David Medina (2-XXXX) Change EM86 to only allow applications to clear the EM bit if a math coprocessor is installed. Changed VNPX to simulate exception 7 instead of interrupt 7 for coprocessor not present exceptions. ═══ 2.48. *59705 BETA:PCAD application gets Trap C after schematic editor piduplicate ═══ name 59705 reference abstract BETA:PCAD application gets Trap C after schematic editor piduplicate state closed severity 3 answer dsai symptom thipe duplicate defects: none. duplicate features: none. notes: BETA: User has reported this problem and I have reproduced it here. The user has PCAD application which comes from CADAM, an IBM subsidiary. This application ran on GA and the service pack, but gets a trap c with the 6.479c driver. -------------------------------------------------------------------- I run PCAD in a Dos Fullscreen session on OS2GA+SP with no problem. On OS2 2.1 Beta, I get a trap 000C starting the schematic editor every time. One small hint; I know it uses DPMI, and I allocate a big chunk of Memory for DMPI in Dos Settings. Two times out of two tries, the TRAP 00C occurred at CS:EIP 0160:fff5f971 A couple of other fields that may be important: ##0160:fff60967 - 000d:9967 60000,9084 048600B4 Internal 4.479 92/12/09 Much more info available on request. Please contact either myself 3-XXXX or Mike Turner 2-XXXX to recreate. Please pass this on to another component if I have indicated the wrong one. Assigning to proper component. Changed VDPX to always insure that the proper host data area is loaded, otherwise we were looking at the incorrect process data. Modified VDPMI to keep track of segment to selector mappings on a per DPMI app basis. Fixed VDHStartHWInt to set and clear the bit in correct copy of flVdmStatus to indicate if we have to restore the previous app's context or not. ═══ 2.49. *59709 Beta: clipboard does not display Metafiles ═══ name 59709 reference abstract Beta: clipboard does not display Metafiles duplicate state closed severity 3 answer lfi symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: When one copies a metafile to the clipboard, the title Bar says Metafile yet render will not display the the Metafile in the clipboard. Document# s61g-XXXX-00 OS/2 2.1 Using the Operating System on page 18-31 states: "When a metafile is copied to the clipboard, it is converted into a bitmap, and then the image is enlarged. Only the center of the image is in the clipboard viewer. This will be fixed in a later release." The metafile is in the clipboard because it can be pasted to another picture viewer, but this document seems to imply that some image should be in the clipbaord viewer when there is none. please contact Lisa Garcia for a detailed description of why I rejected this defect. Apparantly she has identified problems that are not clearly visable at first glance. CMVC:Fixed incorrect AFLD2 field. CMVC: cloned to new defect 60601. ═══ 2.50. *59712 Beta: Selective install of SVGA support ═══ name 59712 reference abstract Beta: Selective install of SVGA support duplicate state closed severity 3 answer chki symptom fnwad duplicate defects: duplicate features: none. notes: BETA: After installing with the primary display as an 8514a using the ATI Ultra/8514a display adapter, selectively install SVGA support for the primary display. The selective install procedure will ask for all the appropriate diskettes then a dialog box with three different resolutions will appear. The Help button on this dialog box does not work and no matter which choice one selects, after shutting down and rebooting the OS/2 logo's will appear and then a black screen without a cursor. System remains in this state. When I originally wrote this defect I identified the ATI Graphics Ultra as a single instance of this problem. Since that time it has come to my attention that a user has identified the exact same black screen after selectively installing SVGA support and rebooting using the Diamon Speedstar 24X with the WD90C31 chipset. The ATI Graphics Ultra Card is a 8514/a card with SVGA support. Dspinstl.exe was recognizing it as an ATI VGA Wonder Card in svga mode. ═══ 2.51. 59721 Beta: Details View does not show details ═══ name 59721 reference abstract Beta: Details View does not show details duplicate state canceled severity 3 answer symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: If one selects a details view on a folder, none of the ═══ 2.52. 59722 Beta: Details View does not show details ═══ name 59722 reference abstract Beta: Details View does not show details duplicate state canceled severity 3 answer invwad symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: If one selects a details view on a folder, none of the Details are shown. A possible side effect of this is sort by last access time real name etc. does not function. If these are unrelated please advise as I will open a different defect for it. Charlene: I am reassigning the CAREY control Defects to your name per Claudette. Thanks - How can we recreate this problem? Also, please, one problem per defect. James Taylor 982-XXXX Simply open up a details view of productivity folder and then open a details view of OS2 System folder. The OS2 system folder has all the fields filled in, last access date/time etc. The productivity folder has only the very first FAXPM. If one goes to a drive all the fields for all the files show up fine there too. It seems to be different in different places. there are not details shown for the program objects because they do not have file system attributes (they only exist in workplace, not in file system) so the details for date, time, etc. don't show because there is no such thing. Open the apps folder off of os2 from drive c and you will see the program files with the attributes shown. Yes this is the same as the sort problem but it's not a problem, it's working as designed. Mindy Pollack ═══ 2.53. 59725 Beta: Installation Regression ═══ name 59725 reference abstract Beta: Installation Regression duplicate state canceled severity 3 answer invsug symptom thipe duplicate defects: none. duplicate features: none. notes: BETA: Trap D on install diskette #1 SYSTEM: 486/50DX Clone (no-name) EISA with 256k Cache Award Bios 32MB Ram Adaptec 1542b with 340MB Maxtor hard drive and Toshiba 3301 CD-ROM Situation: 1) Boot from installation diskette 2) Insert Diskette #1 3) CD-ROM light flashes 4) Trap 000d Trap 000d ERRCD=7d38 ERACC=**** ERLIM=******** EAX=00c8c317 EBX=7d3aff37 ECX=00000000 EDX=fff29246 ESI=7d39ff51 EDI=7d39096e EBP=00004c96 FLG=00013206 CS:EIP=0150:000050f1 CSACC=009b CSLIM=0000dc27 SS:ESP=0030:00004c62 SSACC=1097 SSLIM=00003fff DS=30 DSACC=1097 DSLIM=00003fff CR0=8001001b ES=30 ESACC=1097 ESLIM=00003fff CR2=7d477008 FS=00 FSACC=**** FSLIM=******** GS=00 GSACC=**** GSLIM=******** The system detected an internal processing error at location ##0160:fff60967 -000d:9967 60000,9084 048600b4 Internal Revision 6.479 92/12/09 This is a regression as GA as well as previous BETA installed on this system without any problem. This happens both from the CD install as well as a full diskette install. Trap is the same. The name of the individual is Paul Robertson (508)372-XXXX if a remote debug is needed. The cs:eip 150:50f1 points to the file system, either w_opencreate for debug kernel or dos_opencreate for a retail kernel. Since we donot have the hardware on site I suggest that the user is sent a copy of a post beta build as many changes have gone into 481 and it is impossible to debug back level code. G. Simco ═══ 2.54. 59741 BETA: Quicken for Windows fails in WinOS2 - Borg beta ═══ name 59741 reference abstract BETA: Quicken for Windows fails in WinOS2 - Borg beta duplicate state closed severity 3 answer dsai symptom comp duplicate defects: duplicate features: none. notes: Comment from customer: "Quicken\W runs better in WinOS2 than in MS Windows 3.1" (thought you'd enjoy that...) Problem - Quicken\W has a backup facility that is available from the toolbar or from the FILE item in the Windows menu bar. This backup facility will copy the current file to any user specified target. It will not run under WinOS2, however. Received the error message (the Windows stop sign) that reads, "Unable to open source file." Verified problem by running it under MS Windows 3.1 / Dos 5.0 and it worked OK there. Dan Lee when quicken opens the data file 1st time, it opens with mode AX=3d02h, int 21h. OS2K386 (i21entry.asm) filter it out with AX=3d22h, which is open Read/Write deny_write sharing. In Quicken for Windows, the code of backup option call directly with AX=3d02h, and we turn around issuse AX=3d22h and it fails with AX=0005 (access denied). The reason it fiails is because the file already opened first time with mode deny_write. The correction to this problem is very simple. If the request to open file with mode AX=3d02 (read/write compatibility mode), we should change AX=3d42 (read/write deny none) instead AX=3d22. With deny none it is much more similiar to compatibily mode. The fixed is in i21entry.asm, PATHDSDXCALL label. Joe Luu 26843 ═══ 2.55. 59744 beta: soundblaster not working - missing audiof.dll ═══ name 59744 reference abstract beta: soundblaster not working - missing audiof.dll duplicate state canceled severity 2 answer invnos symptom appf duplicate defects: none. duplicate features: none. notes: During installation a dll needed for audio is not copied to the \mmos2\dll directory. This file audiof.dll is needed for the soundblaster card to work under WIN/OS2. For technical clarification contact Chris Dinallo, MMPM/2 Audio Subsystems Many internals and ISV's have appended to the OS2BETA forums regarding this issue. Marc Schneiderman BTIG 3-XXXX I spoke with Paul Rogers of the Multimedia group, who agreed that this problem pertains to Multimedia install, not to OS/2 install. To route this to the correct group, you need to change the family to "mme" and the component to "inst". (I tried, but I don't have the authority to write defects against multimedia). ═══ 2.56. *59797 BETA: GpiKerningPairs fails if Kerning Pairs > 512 ═══ name 59797 reference abstract BETA: GpiKerningPairs fails if Kerning Pairs > 512 duplicate state closed severity 3 answer inti symptom fnwad duplicate defects: none. duplicate features: none. notes: If an ATM font is loaded whose font metrics returns more than 512 kerning pairs then a call to GpiKerningPairs fails. This problem applies ONLY to a printer presentation space. This problem exhibits itself with HP, PSCRIPT, and EPSON printer drivers. This problem is exhibited in the DESCRIBE word processor when a printer is associated with a document and the ATM font named Charlomane Reg is used. For additional information contact Marc Schneiderman BTIG 3-XXXX VM ID v$ismarc The thunk for GreGetPairKerningTable did allocate a table for the larger 32-bit version of the Kerning pairs. This caused an overrun of the page when >512 pairs of 8 bytes each was used. ═══ 2.57. *59802 BETA: Delete XGA$DMQS after install finished ═══ name 59802 reference abstract BETA: Delete XGA$DMQS after install finished duplicate state closed severity 3 answer chki symptom use duplicate defects: none. duplicate features: none. notes: Code should be updated to delete XGA$DMQS upon completion of the installation of OS/2 2.0 / Additionally code should delete the empty directories of PSFONTS\PFM and PSFONTS at completion of the installation. Additionally: IF NO games are installed and empty Games folder exists. If Picture Viewer applet is specified to NOT be installed then NO icon appears in the productivity folder BUT PICVIEW.EXE , PICVIEW.DLL and PICVIEW.HLP are installed in the respective directories. ═══ 2.58. *59807 BETA: Crosshatched icon after program terminates ═══ name 59807 reference abstract BETA: Crosshatched icon after program terminates duplicate state closed severity 3 answer dsai symptom vis duplicate defects: duplicate features: none. notes: Icon indicates the application is running after termination. DOS app pointing to .BAT file. Shadow of object in the STARTUP folder. Original on desktop is crosshatched and remains cross- hatched after program terminates. The SHADOW in the startup folder is NOT crosshatched. This doesn't belong in the ICON component.... ICON is supposed to be for problems with icon design, etc. Not for functional problems. Looks like this bug is in Workplace, it is responsible for the crosshatch. David A Kerr ═══ 2.59. 59810 BETA: Quicken for Windows can't find file ═══ name 59810 reference abstract BETA: Quicken for Windows can't find file duplicate state canceled severity 3 answer symptom fnwad duplicate defects: none. duplicate features: none. notes: Using Quicken for Windows V2 and backing up the current file, When backing up the current file error message appears: Cannot find source file, file not backed up This worked correctly with GA + SP. I am cancelling this defect due to it being a duplicate of # 59741. ═══ 2.60. 59812 BETA: Description for WIN_RUNMODE not matching the buttons ═══ name 59812 reference abstract BETA: Description for WIN_RUNMODE not matching the buttons duplicate state canceled severity 3 answer symptom vis duplicate defects: none. duplicate features: none. notes: The description for WIN_RUNMODE does not match the buttons. The description in HELP does not tell when to use 3.1 enhanced versus standard modes. I am cancelling this defect due to it being a duplicate of # 59480. ═══ 2.61. 59813 BETA: Prompt misleading when installing on 486 slc2 ═══ name 59813 reference abstract BETA: Prompt misleading when installing on 486 slc2 duplicate state canceled severity 3 answer invsug symptom vis duplicate defects: none. duplicate features: none. notes: When installing on a 486 slc2 the install prompted for the 'hardware software diskette'. Not knowing what that is - and pressing HELP gives : ' Help request for reference disk '. Prompt message should indicate reference disk in the litereal text. Actually, the prompt is for a Hardware Systems Programs Diskette and there is a good reason for it (according to the Engineering Software group) whose diskette it is and who gave us the label. The intention is that at some point in the not too distant future, they will produce a diskette with the ABIOS for all of the RAM-loaded ABIOS machines. This will not be called a reference diskette, but a Hardware Systems Program Diskette. Cancelling defect. Program working as designed. Future releases will ask for a more 'universal' disk as being developed by engineering. ═══ 2.62. 59815 BETA: True type fonts not working from selective install ═══ name 59815 reference abstract BETA: True type fonts not working from selective install duplicate state canceled severity 3 answer symptom fnwad duplicate defects: none. duplicate features: none. notes: True type fonts do not work for some WIN-OS/2 apps after doing selective install. These fonts do work if installed from a full install. Cancelling defect. More information for recreation needed to determine if it is a winos2 or install problem not updateing the system.ini files. ═══ 2.63. 59816 BETA: After install some files show archive bit 'off' ═══ name 59816 reference abstract BETA: After install some files show archive bit 'off' duplicate state canceled severity 3 answer invunr symptom vis duplicate defects: none. duplicate features: none. notes: After installing 2.1 Beta over GA + SP the following files had the archive bit set to 'off'. All other files had them set 'on'. Files in question: os2\dll\bvhvga.dll 12-9-92 5:51p os2\dll\sysmono.fon 12-9-92 7:47p os2\dll\ibmdev32.dll 12-9-92 8:42p The archive bit is set when the file is created (built) and is not touched by the install program. The build group needs to insure that this bit is set on all files before packing them on the diskettes. The files in question have the archive attribute set to on in 6.486 (Martin Bernas B&I 9826910) ═══ 2.64. 59817 BETA: Select printer(s) in printer install only accepts lasduplicate ═══ name 59817 reference abstract BETA: Select printer(s) in printer install only accepts lasduplicate state canceled severity 3 answer intit symptom fnwad duplicate defects: none. duplicate features: none. notes: During Printer installation at the top of the submenu it says "Select printer(s)" - the (s) implying more than one - However it accepts only the last entry. This will require a change to MRI, which is not accepting defects with severity of 4. Since this is a minor problem this is being TRSed to the next release. Cancelling defect. ═══ 2.65. 59818 BETA: omission of uninstall / sys information tool / etc ═══ name 59818 reference abstract BETA: omission of uninstall / sys information tool / etc duplicate state canceled severity 3 answer invsug symptom fnwad duplicate defects: none. duplicate features: none. notes: Many of the features of 2.00.1 such as uninstall, system information and others have been omitted in the beta release. Inclusion of these tools is important in the GA release of the 2.1 These utilities are part of the preload differentiation code, which is the current plan does not provide for releasing these utilities with the base system, a feature with appropriate management approvals will be required in order to implement this suggestion. Cancelling defect. There are no current plans for the incorporation of the requested utilities found in 2.00.1 into the GA release of 2.1 at this time. ═══ 2.66. 59820 BETA: DMPC.EXE required to install LAN Redirector ═══ name 59820 reference abstract BETA: DMPC.EXE required to install LAN Redirector duplicate state canceled severity 3 answer invwad symptom fnwad duplicate defects: none. duplicate features: none. notes: LOGON to Lan Server failed because LAN Redirector requires that DMPC.EXE be installed. This module is installed ONLY if the option of "INSTALLATION AID" is selected, and no DOC or error message is given to explain that to the user. Austin had agreed to highlight this situation in the LAN and CM guides. It is a simple matter to go through the Selective Install to add on the Installation Aid and should be outlined in the dependent software's documentation. This is not an OS/2 base problem. Cancelling defect. This is not a base 2.0 problem. Referenced problem to the Austin Lan people. ═══ 2.67. *59825 BETA:System Setup Not Changing Windows Resolution ═══ name 59825 reference abstract BETA:System Setup Not Changing Windows Resolution duplicate state closed severity 3 answer bldi symptom fnwad duplicate defects: duplicate features: none. notes: Running a mod 95 with XGA, went to system setup and changed the video mode for OS/2 desktop to 640x480x256. Shut down, came backup and started Windows on the desktop. Windows is in hi-res mode. Following note is from Marc Cohen: It appears that the 2.1 BETA was not installing the WINRES.DLL. This is the DLL that does all of the WinOS2 INI file modifications. If the engine cannot load it, (DosLoadModule) then the INI updates cannot be made. Thus it appears that the problem is an install problem. CORRECTION: Install is not installing WININI.DLL. The file name WINRES.DLL is incorrect. The correct name of the missing file is WINPRF.DLL ═══ 2.68. *59830 Standalone Dump hangs since OS2DUMP swapped out ═══ name 59830 reference abstract Standalone Dump hangs since OS2DUMP swapped out duplicate state closed severity 3 answer asnm symptom thipe duplicate defects: duplicate features: none. notes: DHInitSytemDump gets swapped out causing standalone dump to hang. I have a mod 95 with 12 meg memory. With Borg version .479c I am unable to do a standalone dump. The system hangs at ctl-alt-nl-nl before the screen clears and the diagnostic panel remains dark. We switched to a mod 95 with 20 meg. Dumps ok right after a reboot, but hangs if dump after some activity on the system. It appears that the code in os2dump is getting swapped out. Our analysis: U RASRST This is entry point to standalone dump. The call to DHInitSystemDump is via a dword pointer at 400:832. The call to DHSystemDump is ....................... at 400:836 The contents of 400:832 are 100:778 400:836 100:77e U 100:778 call 4a00:0 retf u 100:77e call 4a00:3 retf. On a newly rebooted mod 95 with 20 meg (4a00:0 is real mode entry point supposedly containing a jump) u %%4a000 jmp %something U %%4a003 jmp %something After doing lots of activity on the mod 95, %%4a000 contained hex 00's It appears that the code at %%4a000 was swapped out..... Fixed in 4.486, tested on 4meg mod 70. Flags set by oldr for pg were wrong. Also modifed floppy to not link with rasint15 to make smaller. Paul. ═══ 2.69. 59839 BETA: VER Command in DOS Session ═══ name 59839 reference abstract BETA: VER Command in DOS Session duplicate state canceled severity 3 answer invwad symptom use duplicate defects: none. duplicate features: none. notes: Undocumented, you can see the "/R" parameter in a DOS session and get the version number (6.479 in this case). This has always been true in OS/2 2.X and never documented. Beta Tester --- Jim D. Stuyck and Associates If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX Cmd.exe is not the correct component for this problem. Cmd.exe is the protect mode command shell. Command.com is the real mode command shell (i.e. VDMs, DOS windows and fullscreens). If you are not sure which component to write a defect against, please call the suspected developer and ask before writing the defect. This was a problem caused by those changing the .H and .INC files. Command.com uses the values in the equates to determine the version number. Problem still exists in 492 driver. If any questions, please contact Sheila McDonald-Rich on 3-XXXX. Thanks This defect needs to be assigned to the correct development component. This is not an ID issue. Vince Spadea ID 2-XXXX The /R is a DOS only parameter. It is not enabled for OS/2 sessions. Vince Spadea ID 2-XXXX The remaining problem is the fact that this parameter in not documented in the online command reference. If you have any question, please contact me on 3-XXXX. Thanks, Sheila McDOnald-Rich ═══ 2.70. 59840 BETA: FORMAT Command Missing /Q Option ═══ name 59840 reference abstract BETA: FORMAT Command Missing /Q Option duplicate state canceled severity 3 answer invsug symptom use duplicate defects: none. duplicate features: none. notes: The DOS sessions are supposed to duplicate the functions available in DOS 5.0. The "/Q" (quick) option of FORMAT has never been in OS/2 2.X but is available in DOS 5.0. Beta Tester --- Jim D. Stuyck and Associates If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX The format utility is not in the cmd.exe component. If you are not sure which component to write a defect against, please call the developer and ask before writing the defect. It is the opener's responsibility to specify the correct component. Since the current plan does not provide for this feature in OS/2 2.X, a feature (DCR) with appropriate management approvals will be required in order to implement this suggestion. Adding such new functionality at this time would impact documentation and MRI; such changes need to be minimized at this time due to NLS considerations. For questions, contact the component owner, Pete Rodriguez, at 2-XXXX. Forwarded Suggestion to Planning ═══ 2.71. 59842 BETA: DIR /S - Incorrect Output ═══ name 59842 reference abstract BETA: DIR /S - Incorrect Output duplicate state canceled severity 3 answer valunr symptom use duplicate defects: none. duplicate features: none. notes: Using the "/S" parameter in any OS/2 session always gives an incorrect grand total byte count in "K" -- VERY incorrect. This is not true if you use "/S" in a DOS session. Beta Tester --- Jim D. Stuyck and Associates If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX This defect has already been written. Please review open defects for a component before writing a new defect. The original problem is fixed in that the "K" is gone. However the # of files and the # of bytes used are not the same value in an OS/2 session as they are in a DOS session. Thanks Sheila McDOnald-Rich CMD.EXE computes the number of 'free bytes' based on data returned by DOSQFSINFO (file system). Similarly, COMMAND.COM computes its free bytes value from data returned by w_Get_Drive_Freespace. w_Get_Drive_Freespace gets its information from the same source as DOSQFSINFO, but it rounds off the numbers before returning to COMMAND.COM This only happens when the drive (where DIR was issued) is other than FAT. Screen Team Approved The OS/2 session for Dir /s returns the correct information for # bytes free. The dos session for Dir /s can return a lower number for # bytes free than the OS/2 session if the total clusters on the disk is > 64k. The reason for this is the # of bytes free is calculated by the following formula # bytes free = free clusters * sectors per cluster * bytes per sector If total clusters is > 64K then the code modifies free clusters, total clusters, and sectors per cluster to make these values fit into words. DOS can only have these values in words. Here is an example to demonstrate the rounding error that can occur in Dos. From w_get_drive_freespace which command.com calls to get the # of bytes free, the following values are returned by QFsInfo free clusters = 0000dbbf DD totals clusters = 00081fe0 DD sectors per cluster = 00000001 DD bytes per sector = 0200 DW After the rounding has occurred to fit these values from double words into words .DOS gets the following values free clusters = dbb total = 81fe sectors per cluster = 0010 bytes per sector = 0200 the # bytes free = dbb * 10 * 200 = 28794880 notice the f from dbbf was lost in this conversion, so the DOS session show 7680 bytes less for the # bytes free than the OS/2 session .9375 * 16 * 512 = 7680 The OS/2 Dir /s on this same HPFS drive shows the correct # bytes free 28802560 = dbbf * 1 * 200 Since DOS can only have these values that it uses to calculate in words a rounding can occur. I am returning this defect PRS, since the code can not be fix to stop this problem from happening. Steve Best 3-XXXX The original problem with the total byte count in "K" has been corrected. However, the number of files, bytes used and bytes free do not match in a dos versus an OS/2 session. If any questions, please contact Sheila McDonald-Rich on 3-XXXX. Thanks I talked to Shelia and she did not see this problem is the most current build. ═══ 2.72. 59852 BETA: PULSE Mispositioned ═══ name 59852 reference abstract BETA: PULSE Mispositioned duplicate state canceled severity 3 answer invsug symptom use duplicate defects: none. duplicate features: none. notes: In the list of applets and games that you can select during installation PULSE is listed among the games rather than among "real programs." This could easily lead to PULSE's exclusion, being considered a game. Beta Tester --- Jim D. Stuyck and Associates If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX This is an installation problem, not a problem with PULSE. This is a question of interpretation - Selective Install lists Pulse in the "Tools and Games" dialog box, which contains both tools and games. Since the tools are listed mostly at the top of the box, and the games are mostly at the bottom, I can understand how one might think Pulse is a game. However, there is nothing in the box to differentiate tools from games, so the fact that Pulse is among the games is not a defect. If for some reason Pulse is omitted during the initial install, it is always possible to re-install it using Selective Install. Changing the location of Pulse in the dialog box is an MRI impact, which we need to minimize due to NLS translation considerations. We do not believe there is sufficient justification for making such a change in this release. Forwarded Suggestion to Planning ═══ 2.73. *59853 BETA: Selective Install Byte Count Off ═══ name 59853 reference abstract BETA: Selective Install Byte Count Off duplicate state closed severity 3 answer chkm symptom use duplicate defects: none. duplicate features: none. notes: The "bytes required" begins with a value of 1,101,004 bytes "needed." That's not correct -- should be zero. Beta Tester --- Jim D. Stuyck and Associates If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX The byte count displayed is now adjusted depending upon whether or not a full install is being carried out. If a selective install is being done, then the byte count will now read zero. ═══ 2.74. *59858 BETA: AOX MM486 - can not install ═══ name 59858 reference abstract BETA: AOX MM486 - can not install duplicate state closed severity 2 answer asni symptom use duplicate defects: duplicate features: none. notes: I have a model 60 with an AOX MM486 board with 16MB on the AOX card and AOX microcode level 2.20. I also have an IBM SCSI adapter with Cache with two attached hard drives. The original mode 60 hard drive and controller have been removed. When I attempt to boot off of the 2.1 beta "Install" disk it briefly reads from the floppy, the drive light goes out, the underline cursor is at the top of my display and then nada! The install diskette is good (a: it was originally created on the drive in my machine, and b: I used it to install beta 2.1 on a model 80.) I should mention that I also had the same problem with the install diskette for the service pack. I got around the problem by using my original 2.0 GA install diskette to boot and then continuing with the SP diskette 1. I tried this with the beta 2.1 and it doesn't work any longer. So I seem to be OUT OF THE WATER! Another note, on the advice of AOX support personnel I changed the memory option on the AOX card from "UNMAPPED <16MB" to "REPORT MICROMASTER". With the card set this way, I can boot off of the install diskette. I get the prompt screen for diskette 1. It started reading diskette 1 ( I can hear the drive "grinding" away for awhile). Then it eventually just hangs with the floppy drive light lit. Beta Tester --- Jeff Klein -- Federal Reserve Bank NY If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX Paul, this is the defect to fill in the "486 AOX no boot" problem number. Fixed in first build after 6.490. Arena info array not correct. Paul. The user also needs an update from AOX to the 2.28 level of Microdcode. ═══ 2.75. 59860 BETA: Suggestion to set Animation Default do Disable ═══ name 59860 reference abstract BETA: Suggestion to set Animation Default do Disable duplicate state canceled severity 4 answer invsug symptom use duplicate defects: none. duplicate features: none. notes: The customer has suggested that the default for animation be set to disabled. He considers it cute but a burder that irritates him. "It makes OS/2 seem slow right out of the box." It is also the single most thing that most irritates him about the product. Beta Tester --- Jim D. Stuyck If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX We set animation on for new users to understand where windows come from and how to get them back. We configure it so advanced users can easily turn it off. --Mindy Forwarded Suggestion to Planning ═══ 2.76. 59861 BETA: Suggestion to Remove Drive A from the Desktop ═══ name 59861 reference abstract BETA: Suggestion to Remove Drive A from the Desktop duplicate state canceled severity 4 answer invsug symptom use duplicate defects: none. duplicate features: none. notes: The customer has suggested that the Drive A be removed from the desk- top. Beta Tester --- Jim D. Stuyck If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX Drive A is on the desktop for users who haven't necessarily found the drives folder yet and it also is very useful. If you don't like it, shred it. --Mindy Forwarded Suggestion to Planning ═══ 2.77. 59862 BETA: Suggestion to Automatically Bypass during install ═══ name 59862 reference abstract BETA: Suggestion to Automatically Bypass during install duplicate state canceled severity 4 answer intmp symptom use duplicate defects: none. duplicate features: none. notes: The customer has suggested that the two programs:BASIC.COM and BASICA>COM are automatically bypassed by the installed process on machines without the IBM ROM. The current method issues a warning message. Beta Tester --- Jim D. Stuyck If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX This is a good suggestion, unfortunately, neither I, nor any of the people I have asked know how to determine if the machine has IBM BIOS in protect mode and install only runs in protect mode. If there were a way to determine the BIOS, this would be an easy way to save 100K on OEM machines and I'd be happy to code it. Forwarded Suggestion to Planning ═══ 2.78. 59864 BETA: EXIT_VDM Not Documented ═══ name 59864 reference abstract BETA: EXIT_VDM Not Documented duplicate state verify severity 3 answer chkm symptom use duplicate defects: none. duplicate features: none. notes: The online Command Reference does not have an entry for EXIT_VDM. Beta Tester --- Jim D. Stuyck If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX The cmd.exe component is not the default component for all problems having to do with command prompts and what you type at them. If you are unsure of the correct component please contact the developer. This defect contains insuffcient information as to the exact nature of the problem, but I think you probably want to send it to ID. This Command needs to be documented in the On-line Command Reference. George Fulk will provide the technical explanation for this command. Information about EXIT_VDM will be documented in the Command Reference in the section about using a specific version of DOS in a VDM session. George Fulk has provided the technical input. Vince Spadea ID 2-XXXX Screen Team Approved ═══ 2.79. *59871 BETA: Veiwing a meta file with 123 for OS/2 (32 bit) clip tduplicate ═══ name 59871 reference abstract BETA: Veiwing a meta file with 123 for OS/2 (32 bit) clip tduplicate state closed severity 3 answer lfi symptom vis duplicate defects: none. duplicate features: none. notes: System: Mod 95 Display: XGA Memory: 12 Meg OS/2 Version: Borg Beta Application: Lotus 123 for OS/2 32 bit version. The problem is easy to recreate by doing the following steps: 1. Bring up 123 for OS/2 32 bit application. 2. Select Utility from the menu bar. 3. Select Draw from Utility menu. 4. Select Graph from Draw menu. You will see that word "Title" is clipped. The longer the this word is the more clipping takes place. This is a critical issue for Lotus. If this is not fixed, Lotus will not certify on XGA. I spoke to Joe Celi and showed him the problem. He says it is a engine problem. I have the Lotus application for the developer that needs it. Bob Gonzalez 982-XXXX Had phone call with Lotus Development and discussed the problem. With new information on the problem, development will investigate furher. Bob Gonzalez 982-XXXX Corrections made for font attributes into metafiles and matching fonts in non-device units. Screen Team Approved ═══ 2.80. 59874 Beta: Closing Winos2 with Orchid PRO II ═══ name 59874 reference abstract Beta: Closing Winos2 with Orchid PRO II duplicate state canceled severity 3 answer invnos symptom prf duplicate defects: none. duplicate features: none. notes: BETA: Using an Orchid Prodesigner II in SVGA modes 800X600X256 and higher, when closing winos2 the screen goes black. There is an audible "snap" from the monitor( NEC Multisync 4A ) and then a strobing "flicker". Finally when the desktop comes back, it "wavers" for a moment before stablizing. This is something that is very hard to describe, but is rather annoying to view. User's have complained and I'm not sure if the drivers are actually working incorrectly or this is due to the monitor taking a long time to switch modes. The flicker and snap accur when the hardware is switching modes and syncing. It is not caused by OS/2. The same thing would occur switching between these modes under DOS. If further explanation is required, contact Tatchi Lay. Duane Wood 2-XXXX ═══ 2.81. 59875 BETA: Quit SD386 debug in full screen hangs session ═══ name 59875 reference abstract BETA: Quit SD386 debug in full screen hangs session duplicate state canceled severity 3 answer valunr symptom thipe duplicate defects: none. duplicate features: none. notes: From: CLEMENTS--LEXVMK Date and time 01/06/93 16:12:51 ========================================================================= Date: 6 January 93, 15:54:25 EST From: CLEMENTS at LEXVMK To: GONZALEZ at BCRVMPC1 Quitting the SD386(version 1.02) debugger in a full screen session hangs the session. The session cannot be closed from the desktop and the system has to be rebooted before the session can be removed. To reproduce the problem is quite simple. All you have to do is bring any application up under the debugger and then try to quit out of the debugger. Some more clues: 1. The problem does NOT occur on any service pack machines that I've tried. 2. The problem does NOT occur in a window on OS/2 2.1. Only full screen. 3. I've traced the problem down a bit. When the debugger quits, one of the functions that gets called is MouClose(). If I build a debugger without the MouClose(), the problem goes away. So, the problem appears to be related to the mouse subsystem. If you need a test case then let me know, or you can also see Arthur Roberts. He has SD386 1.02 ( this is the version that uses the mouse ) and plenty of test cases that we've sent for other problems. Thanks, Joe Clements In trying to reproduce this scenario in the latest 2.1 driver, sd386.exe was used in a os/2 full screen environment and upon exiting, the system went on its merry way without any problems. Joe Clements was contacted and aggreed that if the problem was now unreproducible, the defect should be returned. Gordon Webster T/L 982-XXXX ═══ 2.82. *59893 BETA:XGA PM Driver is Returning Invalid Data ═══ name 59893 reference abstract BETA:XGA PM Driver is Returning Invalid Data duplicate state closed severity 3 answer dsam symptom appf duplicate defects: none. duplicate features: none. notes: XGAPM Driver is returning invalid data when GREGETBITMAPBITS is called using a NULL handle. The XGA PM driver was not handling a special case in the GetBitmapBits DDI. If the engine passes a NULL pointer to the bitmap bits storage area, the PM driver should simply update the color table information and return zero. Joe Celi 1/15/92 ═══ 2.83. 59894 BETA: Designer 3.1 GP faults on LAN ═══ name 59894 reference abstract BETA: Designer 3.1 GP faults on LAN duplicate state canceled severity 3 answer invextsw symptom thipe duplicate defects: none. duplicate features: none. notes: BETA: Designer 3.1 running seamlessly System: PS/2 Ultimedia M57SLC with 486slc2 upgrade XGA adapter with 8514 monitor 16mb ram, 387sx-33Mhz co-processor video resolution 1024X768 Problem: Designer caused a general protection fault in module GDI.EXE at 0016:15ae This occours when designer is running off a Novell server. The customer simply clicks with his mouse button anywhere outside the seamless window. Full screen does not have any problems. This does not occour on GA or GA+SP. Level 3 change team. Defect 59894 - Designer 3.1 GP Faults on LAN As per conversation with Paul Luetje (Micrografx Technical Support) at (214) 234-XXXX, Designer 3.1 PM was a limited BETA product designed with partial functionality on OS/2 1.3 ONLY. Micrografx does not offer or support Designer 3.1 PM. Micrografx does currently offers a Designer 3.1 for OS/2 2.0, which is supported by Micrografx. If you have any questions, please feel free to contact me at 443-XXXX. Defect 59894 - Designer 3.1 GP Faults on LAN I also tested Micrografx Designer 3.1 for Windows on OS/2 6_488. I was unable to reproduce a GP Fault while running the application seamlessly. If you have any questions, please feel free to contact me at 443-XXXX. ═══ 2.84. 59895 BETA:Selective install loses app win settings ═══ name 59895 reference abstract BETA:Selective install loses app win settings duplicate state canceled severity 3 answer valunr symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: User reported this error and I reproduced the problem here- I made the Calc program that we ship in winos2 a separate program. When I look at the settings, win-os2 full screen and windows are viable choices. I then do a selective install and change the keyboard from US to UK. When the system comes up after the boot, and I look at the Calc settings, the win-os2 full screen and windows choices are greyed. The user claims when he clicked on win-os2 settings and came back it said dos settings (I coudn't reproduce that part). The user claimed when he tried to run his app, he got a DOS WINDOW and a message like "MS Windows required to run ...." (I didn't see that either, but I don't know what app he was using.) Selective install again of the win-os2 support, restores the settings on the notebook for the app. Bob, this problem has been fixed since the beta went out. The reinstal ini file that you sent me confirmed it, thanks. ═══ 2.85. 59899 BETA:Desktop Tree View changes font for all desktop icons ═══ name 59899 reference abstract BETA:Desktop Tree View changes font for all desktop icons duplicate defect 59304 state canceled severity 3 answer duplicate symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: User has reported and I have reproduced this problem. When you open a desktop tree view, the font for all desktop icons except the minimised ones is changed to a bigger font (14-system) which is used in the Tree View. The new font stays after the tree view is closed. this is the same problem as 59304. When you set the font on a folder, or open a folder, it changes the font in the wrong views. In your case it was tree view, in their case it was settings. Same problem. --Mindy Pollack ═══ 2.86. 59905 BETA: internal processing errors at... ═══ name 59905 reference abstract BETA: internal processing errors at... duplicate state canceled severity 3 answer invdup symptom comp duplicate defects: none. duplicate features: none. notes: Two users have reported internal processing errors with similar register values. I *think* the guilty routine is _DOSFORCEDELETE, but I'm posting the register information I used to find the routine in case I've gotten it all wrong. IPE at location ##0160:fff60967 - 000d:9967 60000,9084 CS:EIP 0148:00000E24 CSLIM = 0000854A ERRCD=0cb8 ERACC=**** ERRLIM ******** EAX=7d250104 EBX=00000482 ECX=ff080000 EDX=ffe20cb8 ESI=7d130482 EDI=00000498 EBP=00005e3a FLAG=00012202 CS:EIP 0148:00000e24 CSACC=009b CSLIM=0000854a SS:ESP=0030:00005e30 SSACC=1097 SSLIM=00004f4f DS=0400 DSACC=0093 DSLIM=000070c0 CR0=fffffffd ES=0158 ESACC=c093 ESLIM=ffffffff CR2=0011f000 FS=0000 GS=0000 FSACC, FSLIM, GSACC, GSLIM, all equal "*"s. This is happening on a PS2 model 80. I have not reproduced it on my model 95. One person reported that it happened frequently while installing and using LAN Server 3.0. The other did not report what software he was running at the time. Dan Lee 2-XXXX Also, (in case it matters...) both users were using 8514 cards. This is a duplicate of 59595. There was a fix that did not make the beta that is pertinent to this problem. G. Simco ═══ 2.87. 59910 BETA:Clipboard and DDE section (18-28 to 18-32) not updatedduplicate ═══ name 59910 reference abstract BETA:Clipboard and DDE section (18-28 to 18-32) not updatedduplicate state canceled severity 3 answer docmt symptom vis duplicate defects: none. duplicate features: none. notes: BETA: The section of the 2.1 Using the Operating System that deals with clipboard and DDE (chapter 18, pages 18-28 to 18-32) are not valid for 2.1. They describe how 2.0 works. Tony Thomas found the discrepancy. We checked with Marian Lindsey whether our copy of the book was down-level and it looks like it wasn't. We checked with Ayo Anise to confirm that the info we were citing was incorrect. Ayo told us it was wrong and he had submitted corrections back in November. BTIG team leader confirmed that a defect should be written for the problem. This section of the Using book will be corrected by January 22, 1993. Screen Team Approved ═══ 2.88. 59939 BETA: The DIR /S command produces incorrect statistics. If duplicate ═══ name 59939 reference abstract BETA: The DIR /S command produces incorrect statistics. If duplicate state canceled severity 3 answer invdup symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: The DIR /S command produces incorrect statistics. If you do a DIR /S, the bytes used entry has a "K" tagged onto the end of it. This is incorrect. This is a duplicate of 59550. ═══ 2.89. 59940 BETA: Pointer initialization in Desktop setting pages is Baduplicate ═══ name 59940 reference abstract BETA: Pointer initialization in Desktop setting pages is Baduplicate state canceled severity 3 answer dsait symptom fnwad duplicate defects: none. duplicate features: none. notes: After openening the Settings for the Desktop, Move pointer to + page advance button, as you flip through pages the 3rd page of FILE tab and 1st page of GENERAL tab show pointer as text insert bar until moved off page advance button. ONLY on first pass through the pages is this found. To recreate close settings and reopen again. 'Display' is not an actual component.... Sending over to WPSH.. It could actually be a control problem, though.... This is definitely notebook. Important enough for BORG? doubt it. This is a minor problem. We will fix it in a future release, as it is too late in the product cycle to be risking major code effort to fix nits. James Taylor 982-XXXX This needs to be in the next release. It is minor but it is in the initialization phase when it errors. cancel due to lack of priority. ═══ 2.90. 59954 BETA: Use of DELDIR in Config.sys causes TRAP's ═══ name 59954 reference abstract BETA: Use of DELDIR in Config.sys causes TRAP's duplicate state canceled severity 3 answer invinf symptom fnwad duplicate defects: none. duplicate features: none. notes: When the DELDIR option is unremarked in the config.sys file, the user gets frequent trap D, E's mostly seen using DEL *.XXX or DEL XXX.* and also using the Icon editor. DOS and OS/2 version affected. This is a duplicate of defect 57983 which was not built in the beta. The submitters scenario was test on build 481 without a problem. G. Simco This defect was reopened without any text that described why the problem still exists. The submitter must show where and why it is a problem on an r206 driver of 481 or later. G. Simco Cancel due to Problem fix in 6.486 ═══ 2.91. 59977 BETA: Example buttons in INF files not CUA compliant. ═══ name 59977 reference abstract BETA: Example buttons in INF files not CUA compliant. duplicate state canceled severity 3 answer invinf symptom comp duplicate defects: none. duplicate features: none. notes: Using the keyboard in Command Reference, TAB to example then Enter to Open. Using a Mouse you click ONCE to highlight the examples box, then it takes TWO clicks to open examples when it should be ONE click. We are confused as to why you believe this to be a container problem. It sounds to us like a problem with the help. James Taylor 982-XXXX This should go to component IPF, as it is a graphic which looks like a PUSH BUTTON and as such is not forced to be CUA compliant. The button is a graphic interface NOT a Push Button and thus does not have to conform to CUA standards. ═══ 2.92. *59978 BETA: Selective install deinstalls fonts ═══ name 59978 reference abstract BETA: Selective install deinstalls fonts duplicate state closed severity 3 answer chki symptom fnwad duplicate defects: none. duplicate features: none. notes: Selective Install will de-isntall fonts except Courier and Sytem Proprortional if the selective install was done from a remote drive. Debbe Deutsch is aware of the problem. The problem was that during the selective install the semaphore for the makeini's was cleared. This should only be cleared for a full install. The selective install should execute prfwriteprofile string to the appropriate ini files, and the icons associated with the code installed will be generated at reboot time. ═══ 2.93. *59979 BETA: MODE COMx information contradicts itself in same windduplicate ═══ name 59979 reference abstract BETA: MODE COMx information contradicts itself in same windduplicate state closed severity 4 answer doci symptom vis duplicate defects: none. duplicate features: none. notes: In the Command Reference, MODE COMx has three inconsistencies in one of the examples the TO=ON enables write infinite processing, which is contrary to info found if a double click on TO parameter on main screen which is correct. The OCTS is backwards and duplicated when you double click on the OCTS parameter from main screen and scroll thru window there is the contradiction in same window. This hardly looks like a cntr problem :-( James Taylor 982-XXXX Changed phase found to be " beta ". Screen Team Approved ═══ 2.94. 59981 BETA: Errors in grammar and punctuation. ═══ name 59981 reference abstract BETA: Errors in grammar and punctuation. duplicate state canceled severity 4 answer invwad symptom vis duplicate defects: none. duplicate features: none. notes: In the Customizing Your Desktop, "Desktop:description" should be "Desktop, Description" In the Doing Everyday Tasks, under "Working with Objects", "Objects:types of" should be "Objects, types of" In the Using the Help Facility, under "Related Information" "Help Menu Bar Choice." should have trailing period removed. The title "desktop:description" is written to correspond to the exact title of the panel it links to. This reasoning is applied to the other topics mentioned. The titles of panels were apparently not supposed to include commas and the Start Here topics merely reflect the panel titles. Vince Spadea 2-XXXX cancel due to problem is noted correctly, except for period removal from "Help Menu Bar Choice." Changed phase found to be " beta ". ═══ 2.95. 59989 BETA: Using AMIPRO Version 3 File Manager window hangs ═══ name 59989 reference abstract BETA: Using AMIPRO Version 3 File Manager window hangs duplicate state closed severity 3 answer dsai symptom fnwad duplicate defects: duplicate features: none. notes: Open AmiPro version 3 / Click on File Management to open another window / go back and click on AmiPro proper / close AmiPro Upon exiting AmiPro the file management window appers and it can not be gotten rid of. Ok. The problem here is that the Ami Pro File Manager app does not do ShowWindow as most apps do. Instead, it creates its window invisible and then at some later point, calls the DefWindowProc with WM_SETREDRAW. This causes the window's visibility flag to be set, so that upon the next window paint, it will actually get painted to the screen. However, PM is never notified about this flag update. Thus, the black hole for this window never gets made visible. The correction is to call ShowWindow from within the code that sets the flag visible. Mark Rogalski ═══ 2.96. 59990 BETA: AmiPro Version 3 Help opens in the background ═══ name 59990 reference abstract BETA: AmiPro Version 3 Help opens in the background duplicate state canceled severity 3 answer invinf symptom vis duplicate defects: none. duplicate features: none. notes: AmiPro version 3 Click on HELP. It opens in the background and foreground windows must be moved/shut to access it. There is insuffiecient information to reproduce this problem. When I tried to bring up Helps from the drop-down menu they came up correctly. Mark Rogalski ═══ 2.97. 59993 BETA: Startup Folder apps do not minimize as set ═══ name 59993 reference abstract BETA: Startup Folder apps do not minimize as set duplicate state canceled severity 3 answer valunr symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: User has reported this problem and I have recreated. Through the system notebook, set apps to minimize to the desktop. Take any app or object such as OS2 Window and set that to minimize to the viewer or hide. Now if you start the OS2 Window from the desktop, it will minimize to the viewer or hide as directed. If you start the OS2 Window from the Startup Folder, when you try to minimize to the viewer or hide, it will minimize to the desktop instead. Mindy is aware of this problem. I can't reproduce this with my private of 59807. I thought I could before but I must have made a mistake. This is a symptom of the same underlying problem so verify that this defect is unreproduceable in the next build (after 6.481). --Mindy ═══ 2.98. *60004 BETA: Cannot edit title with ALT-left mouse on a dragged obduplicate ═══ name 60004 reference abstract BETA: Cannot edit title with ALT-left mouse on a dragged obduplicate state closed severity 3 answer dsam symptom use duplicate defects: none. duplicate features: none. notes: When dragging an object from the Templates folder onto the desktop and then immediately hitting ALT-left mouse on the object in order to edit the title, one notices that the edit window disappears rapidly and does not allow the user to change the title. A customer reported that in order to change the title he had to wait a couple of seconds before hitting ALT-left mouse on the object. This problem also occurs on GA+SP. ═══ 2.99. 60005 BETA: SELECTIVE INSTALL PROBLEM STILL THERE ═══ name 60005 reference abstract BETA: SELECTIVE INSTALL PROBLEM STILL THERE duplicate state canceled severity 3 answer valunr symptom use duplicate defects: none. duplicate features: none. notes: WHEN YOU DO A "REDIRECTED INSTALL" OVER THE LAN, THEN DO "SELECTIVE INSTALL" TO ADD/RE-ADD SOMETHING, SHUTDOWN, RE-BOOT, ALL YOU HAVE ARE 2 FONTS. THE FILES ARE STILL THERE, BUT THE INI DOES NOT RECOGNIZE THEM UNTIL YOU "EDIT FONT', ADD, ETC TO GET THEM BACK. Beta tester If you have any questions, please contact Sheila McDOnald-Rich on 3-XXXX. This was unreproducible on driver 6.486. ═══ 2.100. 60015 BETA: ICONS should be kept on grid to avoid overlay ═══ name 60015 reference abstract BETA: ICONS should be kept on grid to avoid overlay duplicate state canceled severity 4 answer invsug symptom vis duplicate defects: none. duplicate features: none. notes: There should be a feature to keep the ICONS on grid so it would be easier to line up the ICONS. The desktop should be smart enough to recognize when icons are overlaying each other. IN addition the ability to 'nudge the icons' into position would also be of interest if the 'grid' concept could not be impletmented. This does not belong in the ICON component, probably should be examined by the Workplace team for validity. David A Kerr this is already in the plan for the next release (not 2.1, but beyond). please make suggestion defects sev 4 not sev 3. Cancelling defect. ═══ 2.101. 60016 BETA: Recommend need for msg while desktop builds first timduplicate ═══ name 60016 reference abstract BETA: Recommend need for msg while desktop builds first timduplicate state canceled severity 4 answer intmt symptom use duplicate defects: none. duplicate features: none. notes: The initial installation build of the desktop NEEDS some message displayed saying it is doing something, as users frequently turn off the machine thinking nothing is happening. HIGHLY RECOMMENDED! the wpsh is building the desktop at this time...assigning to wpsh with Caldwell's agreement. THis is a good suggestion but it came a little late. The MRI is frozen and we cannot add any words to the interface. We will definitely consider this for the next release. In the future, please write any suggestion defects as sev. 4 not sev. 3. Mindy Pollack 2-XXXX This NEEDS to be fixed, the MRI can always be changed in the next build after 6.481... Vital to customer satisfaction and for limiting incoming Support Line Calls because of trashed EA's. Steve, MRI for the BORG product is closed. This defect is a suggestion and when I compare it to some of the other APARs and Beta problems I have other problems that are more important. Please see Paul Cheatham and myself before reopening this defect. Thanks! Claudette 2-XXXX This is a BUG and needs to be addressed for the next release, the MRI can be changed sometime. cancel due to priority. ═══ 2.102. 60017 BETA: Compudyne 4DX3 486/33 laptop ═══ name 60017 reference abstract BETA: Compudyne 4DX3 486/33 laptop duplicate defect 58329 state canceled severity 3 answer duplicate symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: Compudyne 4DX3 486/33 laptop with pheonix bios Symptom: Install will not proceed beyond initial Blue IBM Logo screen. This individual( Barry Nance ) is a writer who wrote the book INSIDE OS/2 2.0. His Compudyne notebook ran all other versions of OS/2, but will not install on this machine. The full set of diskettes were created from the CD and DO work correctly on other machines. He can be reached at home ( 203 ) 721-XXXX work (203 ) 728-XXXX Re-routed to OLDR after discussion with Paul Devriendt. Duane Wood 2-XXXX ═══ 2.103. 60018 BETA: format on disks >1024 cyl fails ═══ name 60018 reference abstract BETA: format on disks >1024 cyl fails duplicate state canceled severity 3 answer invsug symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: Error message SYS1274. System: 2 RLL drives with 120MB. Symptom: 1) First drive has a second partition at track 1023. Trying to format this partition results in " ... partition > 4G or starting track > 1023" 2) Second drive cannot be found and is never accessed. even using an initialization string of /V /A:0 /U:1 /GEO( ... ) Irv Spalten had discussed this with Russ Berkoff who indicated that format has some dependencies on the CHS fields of the partition records and on drives with >1024 cylinders the fields are not valid. FAT support for >1024 cylinders was a Borg line item that didn't make it. Jamie Jamison (FDISK) is opposed to this feature, I believe for reasons of compatiblility with existing FAT systems. That answer is unacceptable. Our customers are asking and fully expecting OS/2 to support these large drives with a FAT file system. Per agreement between Jim Magnuson, Paul Cheatham, and Jim Macon, this design change requires a Feature. ═══ 2.104. 60021 BETA: Prodigy full screen causes OS/2 Dsktp to go blank ═══ name 60021 reference abstract BETA: Prodigy full screen causes OS/2 Dsktp to go blank duplicate state canceled severity 3 answer valunr symptom vis duplicate defects: none. duplicate features: none. notes: When using an external modem that IS POWERED OFF. Prodigy running in a full screen dos session causes OS/2 desktop to go blank and lock up machine. To reproduce follow the following steps. 1. Power off external modem 2. Start Prodigy in a full screen dos session 3. Enter user ID and password. 4. Press enter key 5. Using CTRL-ESC return to the OS/2 desktop 6. Wait until desktop goes blank. The only way to recover is a warm boot. The problem happens because of vvshadowdatabgnd in VSVGA.SYS. I cannot tell which OUT DX,AL port is causing it, but if I change the OUTs to NOPs, the desktop does not go blank. I recreated this on a machine in the LAN lab. I need to know what is the SVGA chipset in question and if full SVGA support is enabled, eq SVGADATA.PMI should be present on \OS2 directory. Left a message for Marc Schnieder. The problem in vvShadowDataBgnd was solved and integrated into 6.488. Please, reopen if problem reproducable in 6.488 ═══ 2.105. 60023 BETA: Ami Pro Ver 3 - Floating Smarticons ═══ name 60023 reference abstract BETA: Ami Pro Ver 3 - Floating Smarticons duplicate state canceled severity 3 answer regrt symptom vis duplicate defects: none. duplicate features: none. notes: When running Ami Pro Version 3.0 in a Full screen Win/OS2 session select smart icons. Once Smart Icon Dialog Box appears select FLOATING from the dropdown list box. When the floating smarticon toolbar first appears its border is ORANGE. If you drag the toolbar or resize it the border turns to GREY. This is a defect. But it is very minor and does not prevent or disble any Ami Pro 3.0 functions. Therefore, this problem will be fixed on later s problem will be fixed on later I discussed this problem and the temporary solution with Marc Schneiderman and he agreed. -- Ayo Anise (2-6794) ═══ 2.106. *60033 BETA: HP Laserjet IIP driver shows no cartridges or fonts ═══ name 60033 reference abstract BETA: HP Laserjet IIP driver shows no cartridges or fonts duplicate state closed severity 3 answer chkm symptom vis duplicate defects: none. duplicate features: none. notes: Created a printer object on the desktop using the driver for the HP Laserjet IIP (not pscript). When going into the printer driver page of the settings notebook, double clicked on icon. This brought up Printer Properties page. Related popup dialogs showed NO FONTS or CARTIDGES had been install. The install procedure definitely copied all fonts into \OS2\DLL\HP\PCL directory. This defect explains that there are no fonts or cartridges available on the printer properties after installing from a CD rom, the Borg beta. This error condition is symptomatic of not having any EA's on the file during or after install. I thought there was code in the install process to re-hang the EA's back on the file(s) during install. What do you think, Kim? Marc has access to a machine with a cd-rom and has an can reproduce the problem. He works for Casto, and IBM internals have reported this error on the internal forums. This problem doesn't occur when you install the driver through the print object code. It only occurs when you install the driver at install time (or selective install). We have a special case in the print object for .fnt files that have .eas. We handle them just like the .drv files with .eas. If the guy is installing from a CD rom we re-attach the .fni ea file back onto the .fnt file. I took the liberty of looking at your prnfsfnc.c file and you seem to have a case for re-attaching the .ea file back onto .drv files, but not the .fni/.fnt file case. Thanks, Kim Shepard The check for .fni files was not implemented. ═══ 2.107. 60034 BETA: The welcome screen to the tutorial states: "If duplicate ═══ name 60034 reference abstract BETA: The welcome screen to the tutorial states: "If duplicate state canceled severity 3 answer valunr symptom vis duplicate defects: none. duplicate features: none. notes: BETA: The welcome screen to the tutorial states: "If you are unfamiliar with OS/2 2.0...." This should be changed to "2.0 or 2.1" or "2.X" I don't know why this was returned. valunr, I believe is balid but unreproducable(???). This IS reproducable. Open the information folder, click on turorial and read the first screen. Id indeed does say 2.0. Any questions, please call me. 2-XXXX This was fixed. ═══ 2.108. 60035 BETA: When installing Win-OS/2 printer drivers for a 4019E duplicate ═══ name 60035 reference abstract BETA: When installing Win-OS/2 printer drivers for a 4019E duplicate state canceled severity 3 answer invdld symptom use duplicate defects: none. duplicate features: none. notes: BETA: When installing Win-OS/2 printer drivers for a 4019E LaserPrinter the system prompts you to put Printer Driver Diskette 3 into the drive. You are later given a message that install cannont find IBM40x9.DLL. The file is on diskette 4, not diskette 3. This has been fixed in 6.482 ═══ 2.109. 60038 BETA: listbox doesn't paint properly ═══ name 60038 reference abstract BETA: listbox doesn't paint properly duplicate state canceled severity 3 answer chkit symptom fnwad duplicate defects: duplicate features: none. notes: BETA: Pulled from one of our forums. I think you may recognize the name of the submitter. Listbox Extended Selection repaint problem Listboxes in extended selection mode appear not to paint themselves correctly when the user selects a set of items and then de-selects them one at a time whilst holding down the CTRL key. It appears as though the cursoring rectangles are getting left behind around the unselected items - giving the appearance of spurious horizontal lines between the items in the listbox. WC_LISTBOX control window class is in error - component PMWIN.DLL (the window manager) This is a problem in GA and the latest BORG beta too. would be real nice if this could be fixed before 2.1 ships so that Word Perfect can use listboxes in extended selection mode... Chris Andrew (WordPerfect Corp) >>>>>>>>> LBOXTEST.DEF <<<<<<<<<<<<< NAME lboxtest WINDOWAPI DESCRIPTION 'Listbox testcase' DATA MULTIPLE HEAPSIZE 32768 STACKSIZE 32768 PROTMODE EXPORTS MyWindowProc @1 >>>>>>>>> LBOXTEST.LNK <<<<<<<<<<<<< /NOP /NOI /NOP /NOL /NOD /PM:PM + lboxtest.obj lboxtest.exe lboxtest.map /MAP dde4nbs os2386 /NOD /NOE /NOI /NOP lboxtest.def >>>>>>>>> LBOXTEST.H <<<<<<<<<<<<< #define DebugBox(title, text) WinMessageBox(HWND_DESKTOP,HWND_DESKTOP, (PSZ) tex /* DWFR RRRR ISSS RAAA */ #define OPEN_MODE_DRIVE 0x8000 /* 1--- ---- ---- ---- */ #define OPEN_MODE_WRITE_IS_SYNCHRONOUS 0x4000 /* -1-- ---- ---- ---- */ #define OPEN_MODE_NO_HARD_ERROR 0x2000 /* --1- ---- ---- ---- */ #define OPEN_MODE_HANDLE_IS_PRIVATE 0x0080 /* ---- ---- 1--- ---- */ #define OPEN_MODE_DENY_READ_WRITE 0x0010 /* ---- ---- -001 ---- */ #define OPEN_MODE_DENY_WRITE 0x0020 /* ---- ---- -010 ---- */ #define OPEN_MODE_DENY_READ 0x0030 /* ---- ---- -011 ---- */ #define OPEN_MODE_DENY_NONE 0x0040 /* ---- ---- -100 ---- */ #define OPEN_MODE_READ_ONLY 0x0000 /* ---- ---- ---- -000 */ #define OPEN_MODE_WRITE_ONLY 0x0001 /* ---- ---- ---- -001 */ #define OPEN_MODE_READ_WRITE 0x0002 /* ---- ---- ---- -010 */ /* */ #define OPEN_ACTION_FAIL_IF_EXISTS 0x0000 /* ---- ---- ---- 0000 */ #define OPEN_ACTION_OPEN_IF_EXISTS 0x0001 /* ---- ---- ---- 0001 */ #define OPEN_ACTION_REPLACE_IF_EXISTS 0x0002 /* ---- ---- ---- 0010 */ #define OPEN_ACTION_FAIL_IF_NEW 0x0000 /* ---- ---- 0000 ---- */ #define OPEN_ACTION_CREATE_IF_NEW 0x0010 /* ---- ---- 0001 ---- */ >>>>>>>>> LBOXTEST.C <<<<<<<<<<<<< /****************************** Module Header ******************************\ * Module Name: lboxtest \***************************************************************************/ #define DebugBox(title, text) WinMessageBox(HWND_DESKTOP,HWND_DESKTOP, (PSZ) tex /* DWFR RRRR ISSS RAAA */ #define INCL_WINWINDOWMGR #define INCL_WINMESSAGEMGR #define INCL_WINPROGRAMLIST #define INCL_WINSHELLDATA #define INCL_WININPUT #define INCL_WINSYS #define INCL_WIN #define INCL_GPI #define M_I86L #include < stdio.h> #include #include #include "lboxtest.h" #define ID_MYFRAME 1 MRESULT EXPENTRY MyWindowProc( HWND, ULONG, MPARAM, MPARAM ); int main( void ); #define WM_PPAINT 0x0011 #define WM_PSETFOCUS 0x0012 #define WM_PSYSCOLORCHANGE 0x0013 #define WM_PSIZE 0x0014 #define WM_PACTIVATE 0x0015 #define WM_PCONTROL 0x0016 HSWITCH hGreyItem; HWND hwndWindow; HWND hwndFrame; HAB hab; HMQ hmq; /*----------------------------------------------------------------*/ /* Function: main */ /* */ /* Description: */ /* Main line of program */ /*----------------------------------------------------------------*/ int main( ) { QMSG qmsg; ULONG flCreate; CHARStringН50┘; ULONGi; hab = WinInitialize( 0 ); hmq = WinCreateMsgQueue( hab, 0 ); WinRegisterClass( /* Register Window Class */ hab, /* Anchor block handle */ "MyWindow", /* Window Class name */ MyWindowProc, /* Address of Window Procedure */ CS_SIZEREDRAW, /* Class Style */ 0 /* No extra window words */ ); flCreate = FCF_TITLEBAR | FCF_SYSMENU | FCF_MINMAX | FCF_SIZEBORDER | FCF_MENU; hwndFrame = WinCreateStdWindow( HWND_DESKTOP, /* Desktop Window is parent */ FS_ACCELTABLE, /* Frame styles */ (PVOID)&flCreate, /* Control Data */ WC_LISTBOX, /* Window Class name */ "Listbox Extended Selection bug", /* Window Text */ WS_VISIBLE | LS_EXTENDEDSEL, 0, ID_MYFRAME, /* Window ID */ (PHWND)&hwndWindow /* Client Window handle */ ); WinSetWindowPos( /* Set the size and position of */ hwndFrame, /* the window before showing. */ HWND_TOP, 50, 50, 300, 180, SWP_SIZE | SWP_MOVE | SWP_ZORDER | SWP_ACTIVATE ); WinShowWindow( hwndFrame, TRUE ); /* Make visible. */ DebugBox("Listbox Extended Selection bug","Select and unselect multiple items. S tbox, especially when selecting/desecting adjacent entries"); for (i = 0; i < 50; i++) { sprintf(String,"Test Item No.%d",i); WinSendMsg(hwndWindow,LM_INSERTITEM,(MPARAM)i,(MPARAM)String); } while( WinGetMsg( hab, (PQMSG)&qmsg, (HWND)NULL, 0, 0 ) ) WinDispatchMsg( hab, (PQMSG)&qmsg ); WinDestroyWindow( hwndFrame ); WinDestroyMsgQueue( hmq ); WinTerminate( hab ); return(0); } /*----------------------------------------------------------------*/ /* Function: MyWindowProc */ /* */ /* Description: */ /* Client window procedure for this application */ /*----------------------------------------------------------------*/ MRESULT EXPENTRY MyWindowProc( HWND hwnd, ULONG message, MPARAM mp1, MPARAM mp2) { static USHORT vfs = 0; static UCHAR vscancode = 0; static USHORT vchr = 0; static USHORT vvkey = 0; SWCNTRL swle; PID pid; TID tid; VOID OpenFileHandle(PUSHORT); CHAR StringН1000┘; HPS hps; RECTL rcl; POINTL pointl; BOOL brc; USHORT Count=0; PCHRMSG pChrMsg; switch( message ) { case WM_CHAR: pChrMsg = (PVOID)&mp1; if ( !(pChrMsg->fs & KC_KEYUP) ) { vfs= pChrMsg->fs; vscancode = pChrMsg->scancode; vchr = pChrMsg->chr; vvkey = pChrMsg->vkey; WinInvalidateRect(hwnd,NULL,FALSE); } break; case WM_PAINT: hps = WinBeginPaint( hwnd, NULLHANDLE, &rcl ); /* Clear Background First */ WinQueryWindowRect( hwnd, &rcl ); WinFillRect( hps, &rcl, CLR_RED ); /* Then draw character string */ pointl.x = 10; pointl.y = 90; GpiSetColor( hps, CLR_YELLOW ); GpiSetBackColor( hps, CLR_RED ); GpiSetBackMix( hps, BM_OVERPAINT ); sprintf(String, "Flags = %s%s%s", (vfs & KC_SCANCODE) ? "KC_SCAN" : "", (vfs & KC_CHAR) ? " | KC_CHAR" : "", (vfs & KC_VIRTUALKEY) ? " | KC_VKEY" : "" ); GpiCharStringAt( hps, &pointl, strlen(String), String); pointl.y -= 20; sprintf(String, "Scan Code = 0x%x", vscancode); GpiCharStringAt( hps, &pointl, strlen(String), String); pointl.y -= 20; sprintf(String, "Char Code = 0x%x", vchr); GpiCharStringAt( hps, &pointl, strlen(String), String); pointl.y -= 20; sprintf(String, "Virtual Key = 0x%x", vvkey); GpiCharStringAt( hps, &pointl, strlen(String), String); pointl.y -= 20; WinEndPaint( hps ); break; case WM_CLOSE: WinRemoveSwitchEntry(hGreyItem); WinPostMsg( hwnd, WM_QUIT, 0L, 0L ); /* Cause termination */ break; default: return WinDefWindowProc( hwnd, message, mp1, mp2 ); } return FALSE; } /*----------------------------------------------------------------*/ /* Function: OpenFileHandle */ /* */ /* Description: */ /* Recursively opens a file, incrementing a count variable each */ /* time. Thus when the function returns, count will equal the */ /* number of file handles currently available. */ /*----------------------------------------------------------------*/ VOID OpenFileHandle( PUSHORT pCount) { HFILE Handle; ULONG action = 0; ULONG filesize = 10L; ULONG fileattribute=0; ULONG openflag = 0; ULONG openmode = 0; APIRET rc=0; openflag = OPEN_ACTION_CREATE_IF_NEW | OPEN_ACTION_OPEN_IF_EXISTS ; openmode = OPEN_MODE_DENY_NONE | OPEN_MODE_READ_ONLY ; rc=DosOpen("KBD■" ,&Handle ,&action ,filesize ,fileattribute ,openflag ,openmode ,0L ); if(!rc) { (*pCount)++; OpenFileHandle(pCount); DosClose(Handle); } } >>>>>>>>> LBOXTEST.RC <<<<<<<<<<<<< #include #define idMyFrame 1 MENU idMyFrame PRELOAD BEGIN SUBMENU "Exit", -1 BEGIN MENUITEM "Exit Program", 100, MIS_TEXT MENUITEM "Resume Program", 101, MIS_TEXT END END ACCELTABLE idMyFrame PRELOAD BEGIN "a",1, CHAR, ALT "b",2, CHAR, ALT "c",3, CHAR, ALT END //RCINCLUDE GENERIC.RC >>>>>>>>> MAKEFILE <<<<<<<<<<<<< all: lboxtest.exe lboxtest.exe: lboxtest.obj lboxtest.res link386 @lboxtest.lnk mapsym lboxtest rc lboxtest.res lboxtest.obj: lboxtest.c icc /Ss /Q /c /Ti /Gmdn lboxtest.c lboxtest.res: lboxtest.rc rc -r lboxtest.rc Extra X400 information begins: Originator Name: Christopher Andrew Organisation: CORP As per direction from Claudette, changing component to WPSH so that this shows up under Claudette, not JimmyG. Yes, this is Charlene's component but it is part of pmwin not wpsh. --Mindy Mindy. I agree that this is the PMWIN Component. Charlene is working This problem. BUT just because it is on Charlene's plate doesn't put it on your manager's books. This was her idea, not mine. In the future I believe you and I need to talk on these, or you should talk to Claudette so we don't play defect ping pong. - JimmyG I agree this problem needs to be addressed but as we are trying to close down the product and this is not serious enough to be ship issue, I'm TRSing this to the next release. ═══ 2.110. 60039 BETA: Pulled from one of our forums. I think you may recogduplicate defect 60038 ═══ name 60039 reference abstract BETA: Pulled from one of our forums. I think you may recogduplicate defect 60038 state canceled severity 3 answer duplicate symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: Pulled from one of our forums. I think you may recognize the name of the submitter. Listbox Extended Selection repaint problem Listboxes in extended selection mode appear not to paint themselves correctly when the user selects a set of items and then de-selects them one at a time whilst holding down the CTRL key. It appears as though the cursoring rectangles are getting left behind around the unselected items - giving the appearance of spurious horizontal lines between the items in the listbox. WC_LISTBOX control window class is in error - component PMWIN.DLL (the window manager) This is a problem in GA and the latest BORG beta too. would be real nice if this could be fixed before 2.1 ships so that Word Perfect can use listboxes in extended selection mode... Chris Andrew (WordPerfect Corp) >>>>>>>>> LBOXTEST.DEF <<<<<<<<<<<<< NAME lboxtest WINDOWAPI DESCRIPTION 'Listbox testcase' DATA MULTIPLE HEAPSIZE 32768 STACKSIZE 32768 PROTMODE EXPORTS MyWindowProc @1 >>>>>>>>> LBOXTEST.LNK <<<<<<<<<<<<< /NOP /NOI /NOP /NOL /NOD /PM:PM + lboxtest.obj lboxtest.exe lboxtest.map /MAP dde4nbs os2386 /NOD /NOE /NOI /NOP lboxtest.def >>>>>>>>> LBOXTEST.H <<<<<<<<<<<<< #define DebugBox(title, text) WinMessageBox(HWND_DESKTOP,HWND_DESKTOP, (PSZ) tex /* DWFR RRRR ISSS RAAA */ #define OPEN_MODE_DRIVE 0x8000 /* 1--- ---- ---- ---- */ #define OPEN_MODE_WRITE_IS_SYNCHRONOUS 0x4000 /* -1-- ---- ---- ---- */ #define OPEN_MODE_NO_HARD_ERROR 0x2000 /* --1- ---- ---- ---- */ #define OPEN_MODE_HANDLE_IS_PRIVATE 0x0080 /* ---- ---- 1--- ---- */ #define OPEN_MODE_DENY_READ_WRITE 0x0010 /* ---- ---- -001 ---- */ #define OPEN_MODE_DENY_WRITE 0x0020 /* ---- ---- -010 ---- */ #define OPEN_MODE_DENY_READ 0x0030 /* ---- ---- -011 ---- */ #define OPEN_MODE_DENY_NONE 0x0040 /* ---- ---- -100 ---- */ #define OPEN_MODE_READ_ONLY 0x0000 /* ---- ---- ---- -000 */ #define OPEN_MODE_WRITE_ONLY 0x0001 /* ---- ---- ---- -001 */ #define OPEN_MODE_READ_WRITE 0x0002 /* ---- ---- ---- -010 */ /* */ #define OPEN_ACTION_FAIL_IF_EXISTS 0x0000 /* ---- ---- ---- 0000 */ #define OPEN_ACTION_OPEN_IF_EXISTS 0x0001 /* ---- ---- ---- 0001 */ #define OPEN_ACTION_REPLACE_IF_EXISTS 0x0002 /* ---- ---- ---- 0010 */ #define OPEN_ACTION_FAIL_IF_NEW 0x0000 /* ---- ---- 0000 ---- */ #define OPEN_ACTION_CREATE_IF_NEW 0x0010 /* ---- ---- 0001 ---- */ >>>>>>>>> LBOXTEST.C <<<<<<<<<<<<< /****************************** Module Header ******************************\ * Module Name: lboxtest \***************************************************************************/ #define DebugBox(title, text) WinMessageBox(HWND_DESKTOP,HWND_DESKTOP, (PSZ) tex /* DWFR RRRR ISSS RAAA */ #define INCL_WINWINDOWMGR #define INCL_WINMESSAGEMGR #define INCL_WINPROGRAMLIST #define INCL_WINSHELLDATA #define INCL_WININPUT #define INCL_WINSYS #define INCL_WIN #define INCL_GPI #define M_I86L #include #include #include #include "lboxtest.h" #define ID_MYFRAME 1 MRESULT EXPENTRY MyWindowProc( HWND, ULONG, MPARAM, MPARAM ); int main( void ); #define WM_PPAINT 0x0011 #define WM_PSETFOCUS 0x0012 #define WM_PSYSCOLORCHANGE 0x0013 #define WM_PSIZE 0x0014 #define WM_PACTIVATE 0x0015 #define WM_PCONTROL 0x0016 HSWITCH hGreyItem; HWND hwndWindow; HWND hwndFrame; HAB hab; HMQ hmq; /*----------------------------------------------------------------*/ /* Function: main */ /* */ /* Description: */ /* Main line of program */ /*----------------------------------------------------------------*/ int main( ) { QMSG qmsg; ULONG flCreate; CHARStringН50┘; ULONGi; hab = WinInitialize( 0 ); hmq = WinCreateMsgQueue( hab, 0 ); WinRegisterClass( /* Register Window Class */ hab, /* Anchor block handle */ "MyWindow", /* Window Class name */ MyWindowProc, /* Address of Window Procedure */ CS_SIZEREDRAW, /* Class Style */ 0 /* No extra window words */ ); flCreate = FCF_TITLEBAR | FCF_SYSMENU | FCF_MINMAX | FCF_SIZEBORDER | FCF_MENU; hwndFrame = WinCreateStdWindow( HWND_DESKTOP, /* Desktop Window is parent */ FS_ACCELTABLE, /* Frame styles */ (PVOID)&flCreate, /* Control Data */ WC_LISTBOX, /* Window Class name */ "Listbox Extended Selection bug", /* Window Text */ WS_VISIBLE | LS_EXTENDEDSEL, 0, ID_MYFRAME, /* Window ID */ (PHWND)&hwndWindow /* Client Window handle */ ); WinSetWindowPos( /* Set the size and position of */ hwndFrame, /* the window before showing. */ HWND_TOP, 50, 50, 300, 180, SWP_SIZE | SWP_MOVE | SWP_ZORDER | SWP_ACTIVATE ); WinShowWindow( hwndFrame, TRUE ); /* Make visible. */ DebugBox("Listbox Extended Selection bug","Select and unselect multiple items. S tbox, especially when selecting/desecting adjacent entries"); for (i = 0; i < 50; i++) { sprintf(String,"Test Item No.%d",i); WinSendMsg(hwndWindow,LM_INSERTITEM,(MPARAM)i,(MPARAM)String); } while( WinGetMsg( hab, (PQMSG)&qmsg, (HWND)NULL, 0, 0 ) ) WinDispatchMsg( hab, (PQMSG)&qmsg ); WinDestroyWindow( hwndFrame ); WinDestroyMsgQueue( hmq ); WinTerminate( hab ); return(0); } /*----------------------------------------------------------------*/ /* Function: MyWindowProc */ /* */ /* Description: */ /* Client window procedure for this application */ /*----------------------------------------------------------------*/ MRESULT EXPENTRY MyWindowProc( HWND hwnd, ULONG message, MPARAM mp1, MPARAM mp2) { static USHORT vfs = 0; static UCHAR vscancode = 0; static USHORT vchr = 0; static USHORT vvkey = 0; SWCNTRL swle; PID pid; TID tid; VOID OpenFileHandle(PUSHORT); CHAR StringН1000┘; HPS hps; RECTL rcl; POINTL pointl; BOOL brc; USHORT Count=0; PCHRMSG pChrMsg; switch( message ) { case WM_CHAR: pChrMsg = (PVOID)&mp1; if ( !(pChrMsg->fs & KC_KEYUP) ) { vfs= pChrMsg->fs; vscancode = pChrMsg->scancode; vchr = pChrMsg->chr; vvkey = pChrMsg->vkey; WinInvalidateRect(hwnd,NULL,FALSE); } break; case WM_PAINT: hps = WinBeginPaint( hwnd, NULLHANDLE, &rcl ); /* Clear Background First */ WinQueryWindowRect( hwnd, &rcl ); WinFillRect( hps, &rcl, CLR_RED ); /* Then draw character string */ pointl.x = 10; pointl.y = 90; GpiSetColor( hps, CLR_YELLOW ); GpiSetBackColor( hps, CLR_RED ); GpiSetBackMix( hps, BM_OVERPAINT ); sprintf(String, "Flags = %s%s%s", (vfs & KC_SCANCODE) ? "KC_SCAN" : "", (vfs & KC_CHAR) ? " | KC_CHAR" : "", (vfs & KC_VIRTUALKEY) ? " | KC_VKEY" : "" ); GpiCharStringAt( hps, &pointl, strlen(String), String); pointl.y -= 20; sprintf(String, "Scan Code = 0x%x", vscancode); GpiCharStringAt( hps, &pointl, strlen(String), String); pointl.y -= 20; sprintf(String, "Char Code = 0x%x", vchr); GpiCharStringAt( hps, &pointl, strlen(String), String); pointl.y -= 20; sprintf(String, "Virtual Key = 0x%x", vvkey); GpiCharStringAt( hps, &pointl, strlen(String), String); pointl.y -= 20; WinEndPaint( hps ); break; case WM_CLOSE: WinRemoveSwitchEntry(hGreyItem); WinPostMsg( hwnd, WM_QUIT, 0L, 0L ); /* Cause termination */ break; default: return WinDefWindowProc( hwnd, message, mp1, mp2 ); } return FALSE; } /*----------------------------------------------------------------*/ /* Function: OpenFileHandle */ /* */ /* Description: */ /* Recursively opens a file, incrementing a count variable each */ /* time. Thus when the function returns, count will equal the */ /* number of file handles currently available. */ /*----------------------------------------------------------------*/ VOID OpenFileHandle( PUSHORT pCount) { HFILE Handle; ULONG action = 0; ULONG filesize = 10L; ULONG fileattribute=0; ULONG openflag = 0; ULONG openmode = 0; APIRET rc=0; openflag = OPEN_ACTION_CREATE_IF_NEW | OPEN_ACTION_OPEN_IF_EXISTS ; openmode = OPEN_MODE_DENY_NONE | OPEN_MODE_READ_ONLY ; rc=DosOpen("KBD■" ,&Handle ,&action ,filesize ,fileattribute ,openflag ,openmode ,0L ); if(!rc) { (*pCount)++; OpenFileHandle(pCount); DosClose(Handle); } } >>>>>>>>> LBOXTEST.RC <<<<<<<<<<<<< #include #define idMyFrame 1 MENU idMyFrame PRELOAD BEGIN SUBMENU "Exit", -1 BEGIN MENUITEM "Exit Program", 100, MIS_TEXT MENUITEM "Resume Program", 101, MIS_TEXT END END ACCELTABLE idMyFrame PRELOAD BEGIN "a",1, CHAR, ALT "b",2, CHAR, ALT "c",3, CHAR, ALT END //RCINCLUDE GENERIC.RC >>>>>>>>> MAKEFILE <<<<<<<<<<<<< all: lboxtest.exe lboxtest.exe: lboxtest.obj lboxtest.res link386 @lboxtest.lnk mapsym lboxtest rc lboxtest.res lboxtest.obj: lboxtest.c icc /Ss /Q /c /Ti /Gmdn lboxtest.c lboxtest.res: lboxtest.rc rc -r lboxtest.rc Extra X400 information begins: Originator Name: Christopher Andrew Organisation: CORP ═══ 2.111. 60040 BETA: Can not Boot DOS, if OS/2 Creates Partition over 1024duplicate ═══ name 60040 reference abstract BETA: Can not Boot DOS, if OS/2 Creates Partition over 1024duplicate state tcanceled severity 2 answer invwad symptom use duplicate defects: none. duplicate features: none. notes: I installed OS/2 2.1 beta. Having two 80M MFM drives, I was excited about support of over 1024 cylinders, since my second drive has 1314 cylinders. Therefore, I created a second partition, made of cyls over 1024, 18Meg, using FDISKPM. Worked great. For the first time, OS/2 can manage more than 10 year old limit of 1024 cylinders! The new IBM1s506.ADD does that. Slight problem, however. Once OS/2 partition over 1024 is created, DOS no longer boots. Not only from any hard drive, but any DOS! Even MS DOS 5.0 from a diskette. It does not like that OS/2 partition, whether FAT or HPFS, over 1024 cyls. Upon boot, the drive light comes on and stays on, that's it. Once that partition is deleted, DOS boots. Since OS/2 does not support programs using VCPI (In my case all important ORACLE database), capability of booting DOS is absolutely necessary. In comparison, DOS works fine when that partition is created by Ontrack disk manager, and accesses that partition via DMDRVR.BIN driver, the facto statndard, that comes with DOS 5.0. Apparently OS/2 does not seem to adhere to that convention. A question to IBM: Considering that OS/2 is a better DOS than DOS, the above incompatibility is: - A temporary bug that will be eventually corrected to be in sync with DOS & Ontrack. - It is not a bug, it works as planned, in this case, OS/2 documentation should be updated with a new restriction, that if you create a part- ition over 1024 cylinders, using FDISKPM, you will no longer be able to boot DOS, even from a DOS installation disk. I hope the answer is the first one. Beta tester If you have any questions, please contact Sheila McDOnald-Rich on 3-XXXX. Rerouting to FDISK component. This is working as designed. The restriction of booting from greater than 1024 cylinders is neither an OS/2 or DOS limit- ation but a CBIOS hardware limit. The int 13 cylinder field is rerstricted to 10 bits, 1023 value. Int 13 is required for booting up until the machine is switched to protect mode. Besides booting, when real DOS is running if a program issued an int 13 it would fail. To circumvent around this I would suggest that you repartition the drive, using OS/2 FDISK and make a partition that doesn't exceed the 1023 limit, and then make a logical partition out of the rest. The less than 1023 partition will be fat and can be booted and accessed by everyone. The partition exceeding the 1023 limit will be hpfs and will be supported by OS/2, includuing the dos box. You may want to run Boot Manager or dualboot to facilliate this. Also some machines will may map this same drive geometry so as utilize the entire drive space without exceeding the 1023 limit. WAD ═══ 2.112. 60041 BETA: Comm Errors ═══ name 60041 reference abstract BETA: Comm Errors duplicate state canceled severity 3 answer invinf symptom use duplicate defects: none. duplicate features: none. notes: I am running PMFax 1.16, however the status box gets covered up and I forget to turn receive off. It's just a nuiscance when I do forget. A beta Tester If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX. Forwarded Suggestion to Planning ═══ 2.113. *60046 BETA: PMSEEK does not retain defualt editor ═══ name 60046 reference abstract BETA: PMSEEK does not retain defualt editor duplicate state closed severity 3 answer chkm symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: When using PMSEEK if one selects an editor other than E.EXE as the editor filespec, select options, select Set defaults, close PMSEEK and reopen it. It again defaults to E.EXE as the editor filespec. This is also the way it functions in GA. ═══ 2.114. *60047 REGR: Icon positions not restored in folders, icons lost. ═══ name 60047 reference abstract REGR: Icon positions not restored in folders, icons lost. duplicate state verify severity 3 answer dsai symptom vis duplicate defects: duplicate features: none. notes: Icon positions are not restored when a folder is closed then opened. Also, icons are lost on subsequent opens. (i.e. Open OS/2 system folder and then move an icon, close the folder and reopen, do this several times with several folders and you will see icon misbehavior). Sometimes the system will appear to hang, due to a loop in VacateFolderThread. I've talked to Peter Magid about this. -Sheila 2-XXXX Left off the check to increment the container record - Peter G. Magid, 2-XXXX Crit: Frito Lay; (S. Wuest) ═══ 2.115. 60052 BETA: Include page of desktop settings has no page nums ═══ name 60052 reference abstract BETA: Include page of desktop settings has no page nums duplicate state canceled severity 4 answer intit symptom vis duplicate defects: none. duplicate features: none. notes: The desktop settings tab "INCLUDE" does not have any page numbers. The "page X of X" part of some tabs does not show entire last number. This has been this way since GA. We already have another TRSed defect for the same problem. James Taylor 982-XXXX This will never get fixed unless we submit to it. I have also found a problem in many of the object settings. System settings, Screen TAB has Page 1 of 2 on it and if try to go to page two, you go to the next TAB!. I could give you many such occurances free for the asking....resubmit! This is a valid defect which make users think there is a page 2 when there is not one. It is confusing and troublesome...It NEEDS to be fixed. Steve, I am TRS'ing defect also. We will have this fixed in the next release. Claudette 2-XXXX Cancel due to priority. Needs to be to be added for next release ═══ 2.116. 60053 BETA: Icons in startup don't unhash ═══ name 60053 reference abstract BETA: Icons in startup don't unhash duplicate defect 59807 state canceled severity 3 answer duplicate symptom vis duplicate defects: none. duplicate features: none. notes: BETA: If one puts a program icon, say a Dos window with starting from a diskette image, into the startup folder, shut down and reboot, the program is started. If one then closes the window from the task list, the icon remains hashed although the program is closed and no longer appears in the task list. this is a duplicate and will be fixed in next borg build duplicate 59807 ═══ 2.117. 60062 BETA: EDIT.HLP file needed for QBASIC /EDIT not in \OS2\MDOduplicate ═══ name 60062 reference abstract BETA: EDIT.HLP file needed for QBASIC /EDIT not in \OS2\MDOduplicate state canceled severity 3 answer docmp symptom fnwad duplicate defects: none. duplicate features: none. notes: Using a DOS session, open QBASIC with /EDIT param, get message "File EDIT.HLP not found, Please insert diskette in drive or change options+help path". If you hit enter it does nothing, you MUST cancel to continue using the editor. I spoke to A.Ford, this is not a sys.editor problem. She suggested that I re route this to DOS Emulation DOS5 or WIN/OS-2. It is not clear to me whether the .hlp file even exists on the harddisk. In either case I do not own QBasic. The OS/2 BORG Project Office, 1st/2nd line development mgrs want this defect cancelled. We do not ship the help according to the Product Plan. Please accept/cancel this defect on your end as I cannot do a cancel and this is what I have been asked to do by the people in charge of OS/2. K.Hamill, Project Office. Canceled Per khamill ═══ 2.118. *60063 BETA: File IBM2M57.ADD placed in \OS2 on NON-IBM machines ═══ name 60063 reference abstract BETA: File IBM2M57.ADD placed in \OS2 on NON-IBM machines duplicate state closed severity 2 answer bldm symptom fnwad duplicate defects: none. duplicate features: none. notes: The file IBM2M57.ADD for the multimedia machines is being placed in the \OS2 subdirectory by the installation even on NON-IBM machine types. assign to ins from inst Deleted IBM1M57.ADD at the end of install. Reopen due to reproduced error in 6.488 on 486-33 Micronics Mthrbrd. CMVC:Fixed incorrect AFLD2 field. CMVC: cloned to new defect 61022. ═══ 2.119. 60069 BETA: Minimized OS/2 programs vanish if in startup folder ═══ name 60069 reference abstract BETA: Minimized OS/2 programs vanish if in startup folder duplicate state canceled severity 3 answer invdup symptom vis duplicate defects: none. duplicate features: none. notes: Programs in startup folder that should start minimize at IPL do not. Created shadow's of EPM.EXE and PULSE.EXE and placed in startup folder. Enabled settings for both programs to minimize to the desktop. Started both programs and minimized to the desktop. OK so far. Shutdown machine and pressed CTRL-ALT -DEL. When system came back up the MINIMIZED pulse and epm were nowhere to be found on the desktop. Hit CTRL-ESC to bring up task list. They were both listed as running. Looked into the minimized window viewer and VOILA they were in there! Contact Marc Schneiderman 3-XXXX for more information. this is a dup of 59993. please look through our database before opening defects. sev one is for build breaks and sev 2 is for test case blocks ═══ 2.120. 60079 BETA: Deadlock using LAN Network Mgr ═══ name 60079 reference abstract BETA: Deadlock using LAN Network Mgr duplicate state canceled severity 3 answer invinf symptom comp duplicate defects: none. duplicate features: none. notes: (I'm not certain I've got the right component, but...) Reported by several users on OS2BETA CFORUM and over the hotlines: The 2.1 beta deadlocks when the icon for LAN Network Manager is selected. The steps are pretty simple - at boot time the LAN requester (2.0) and Comm Manager both come up okay. The user then selects the LNM icon and the system deadlocks. The mouse moves, but the nothing responds to it. Keyboard is also dead, and power-off is required. I called the LAN Network Mgr folks in Raleigh, but nobody could tell me whether anyone there had tested it under 2.1. Another note - customer encountered the "deadlock" late in the evening and just went home, leaving the system running. In the morning the clock was running, sort of, but only activating every three or four minutes (minute hand would jump from 8:20 to 8:24 to 8:27 etc), and his pulse screen was up but the red line had disappeared halfway across it. FWIW.... Dan Lee 2-XXXX I don't understand what the user was starting. What is the lan manager icon? what does it look like and what is the exact text? I'm trying to figure out if this is a lan aware shell problem or a lan manager problem. --Mindy It could be either of the above. IBM's LAN Network Manager is a product out of Raleigh. It should not be confused with IBM's LAN Services or with Microsoft's LAN Mgr. It is in version 1.1 now, but customers have been reporting 2.1 deadlocks using versions 1.0 and 1.1 of the IBM Network Manager. When installed, it puts an icon on the desktop. It worked fine under 2.0, but under the 2.1 beta it deadlocks. Dan Lee this problem should be looked at by the lan network manager folks in Raleigh. Mindy Pollack This program worked on 2.0 GA now it does not work on Borg. I find it hard to believe the problem is the program. I am returning this insufficient infomation per my conversation with Paul Cheatham Fix/Work-around documented in 61554. This was nothing at all to do with 61554. This was a kernel problem that has been fixed. Paul. ═══ 2.121. 60080 Hangs on diskette 1 on Compudyne 486 Notebook ═══ name 60080 reference abstract Hangs on diskette 1 on Compudyne 486 Notebook duplicate defect 58329 state canceled severity 3 answer duplicate symptom thipe duplicate defects: none. duplicate features: none. notes: User has installed Borg on other systems successfully, but on the Compudyne 486DX, the install hangs when diskette 1 is called for. The Compudyne has been running the GA level and the 11/29 level of OS/2 without problems. duplicate of 60017 ═══ 2.122. 60081 BETA: Backup program in 2.1 does not work correctly ═══ name 60081 reference abstract BETA: Backup program in 2.1 does not work correctly duplicate state canceled severity 3 answer invwad symptom fnwad duplicate defects: none. duplicate features: none. notes: Backup program provided in Welcome folder correctly backs up everything - Backup program in 2.1 beta does not. Preload backup : 29712 bytes 10/26/92 2.1 backup : 84069 bytes 12/09/92 IN the 2.1 backup issued BACKUP C:\*.* A: /S did not list or appear to back up system files: OS2BOOT,OS2KRNL OS2LDR WP ROOT. SF / Creating Backup and Restore programs from the Welcome Folder and executing BOOTDISK.EXE. This program allowed backing up the entire system, including the OS2KRNL, allowed formatting of the hard disk prior to restore and allowed a complete restore of the system. BACKUP was designed to back up only user files. In order to have BACKUP to provide the same functionality as PRELOAD BACKUP a new feature would have to be introduced. ═══ 2.123. 60084 BETA: Config.Sys Migration & Novell Netware ═══ name 60084 reference abstract BETA: Config.Sys Migration & Novell Netware duplicate state canceled severity 3 answer invunr symptom use duplicate defects: none. duplicate features: none. notes: When I installed the 2.1 beta on an existing 2.0 system with the netware requester installed, some of my network apps stopped working. Symptoms were: - WordPerfect office was not able to lookup my user-related information in it's internal table -- it would prompt me for my three character initials - Netware Resources item was missing from DOS Settings, but VIPX ENABLED was still present. After checking further I found that DEVICE=C:\NETWARE\VSHELL.SYS was missing from my migrated config.sys For reference: After the original install, I ran install from the system folder to re-install VGA support. The orginal config.sys ended: . CODEPAGE=437,850 DEVINFO=KBD,US,C:\OS2\KEYBOARD.DCP The new config.sys ended: . REM DEVICE and RUN statements from your older CONFIG.SYS DEVICE=C:\OS2\MDOS\VVGA.SYS DEVICE=C:\NETWARE\LSL.SYS RUN=C:\NETWARE\DDAEMON.EXE DEVICE=C:\NETWARE\smc8000.sys DEVICE=C:\NETWARE\IPX.SYS DEVICE=C:\NETWARE\SPX.SYS RUN=C:\NETWARE\SPDAEMON.EXE DEVICE=C:\NETWARE\NWREQ.SYS IFS=C:\NETWARE\NWIFS.IFS RUN=C:\NETWARE\NWDAEMON.EXE DEVICE=C:\NETWARE\VIPX.SYS DEVINFO=SCR,VGA,C:\OS2\VIOTBL.DCP SET VIDEO_DEVICES=VIO_VGA SET VIO_VGA=DEVICE(BVHVGA) with VSHELL nowhere to be found. A BETA TESTER If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX. This problem does not occur on the latest driver. The code that caused similar problems was corrected after the Beta freeze and has been incorporated into the post-beta builds. ═══ 2.124. 60098 BETA: CTRL-ALT-NUMLOCK-NUMLOCK does not work. ═══ name 60098 reference abstract BETA: CTRL-ALT-NUMLOCK-NUMLOCK does not work. duplicate defect 59830 state canceled severity 3 answer duplicate symptom use duplicate defects: none. duplicate features: none. notes: You can not initiate a standalone dump from the keyboard. What I got was a desktop pulldown and a dead system. C-A-D was inoperative after this. A BETA TESTER If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX. Fix is in driver 486 ═══ 2.125. *60105 BETA: User has to insert 13 diskettes for selective installduplicate ═══ name 60105 reference abstract BETA: User has to insert 13 diskettes for selective installduplicate state verify severity 2 answer lfm symptom use duplicate defects: none. duplicate features: none. notes: Known problem since GA, many many people voice disatisfaction at this"blatent unefficiency"...Quote from CIS customer! Changed Severity back to 2 per Owner Screen Team Approved ═══ 2.126. 60106 BETA: Scheme palate fails to change color of shadow text ═══ name 60106 reference abstract BETA: Scheme palate fails to change color of shadow text duplicate state canceled severity 4 answer dsamt symptom fnwad duplicate defects: none. duplicate features: none. notes: Draged Scheme Palate onto the desktop. Selected default scheme. Changed color of shadow text then saved. Color of shadow text failed to change. the interface for changing shadow text has always been there but has never worked. We have to close BORG down now and we can't contain this kind of fix. We'll look at it in the next release. Mindy Pollack 2-XXXX ═══ 2.127. 60110 BETA: Need Welcome Folder for 2.1 ═══ name 60110 reference abstract BETA: Need Welcome Folder for 2.1 duplicate state canceled severity 4 answer invsug symptom use duplicate defects: none. duplicate features: none. notes: This is a suggestion: A couple things from the welcome folder which we shipped with 2.00.1 should be included in 2.1 They are Your System folder, and the selective uninstall folder. A BETA TESTER If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX. thanks for the suggestion but we cannot add any objects to the system this late in the cycle. (mri freezes etc.) Suggestion sent to Planning ═══ 2.128. 60116 BETA: 1.x KbdCharin command fails under 2.1 ═══ name 60116 reference abstract BETA: 1.x KbdCharin command fails under 2.1 duplicate state canceled severity 3 answer invwad symptom comp duplicate defects: none. duplicate features: none. notes: (I'm not certain I've got the right component, but...) A 1.x routine that uses "kbd..." functions fails in the beta (it worked okay under 2.0). The developer reported the following scenario: KbdOpen creates a new logical keyboard and returns a handle. The routine then calls KbdCharin with the IOWait parameter set to 0 (which means to wait). If the third parameter to KbdCharin is set to the handle returned by KbdOpen, KbdCharin does *not* wait, but returns immediately. If the default handle (0) is used instead, then KbdCharin functions properly and waits for a character. Dan Lee 2-XXXX MKIO is the proper component for this defect. The call to KbdOpen forces an application to bond the returned keyboard handle by using the API KbdGetFocus before using that handle for any keyboard API calls. Reference the KbdOpen API in the OS/2 1.3 Command Program Programming Reference for further information. Gordon Webster T/L 982-XXXX Okay, I've forwarded the reply to the customer and will close this. Dan Lee ═══ 2.129. 60118 BETA: Nothing happens when double clicking on README ═══ name 60118 reference abstract BETA: Nothing happens when double clicking on README duplicate state canceled severity 3 answer symptom fnwad duplicate defects: none. duplicate features: none. notes: lect "R" and then the installation goes through, but with doubts about the completion of the installation process. This defect is not for ID. Please forward to the appropriate component. The original text for this defect was somehow corrupted, but this is what's happening: The user goes to the Information folder, then double clicks on the README and nothing happens. The user then decides to hit the right mouse on the object to verify that there is an editor associated with it, but the only option available from the "Open" option is "Settings". I was able to see this on our MOD 95, but I also checked other systems where this is not reproducible. I noticed that the README icon is different in those installations where it does works than it those where it doesn't work. Returning this as unreproducible per discussion with Carmen Hernandez. The problem is not seen on drivers .486 and .488. ═══ 2.130. 60122 BETA: MODE co80,40 cmd breaks ═══ name 60122 reference abstract BETA: MODE co80,40 cmd breaks duplicate state canceled severity 3 answer invunr symptom comp duplicate defects: none. duplicate features: none. notes: Customer installed OS2 beta with display drivers set to SVGA; using an Everex Viewpoint NI card (Tseng based). From an OS2 command prompt, the command MODE CO80,40 gives them a register dump. They didn't append the registers (I've asked for them but they have not replied yet.) Dan Lee More info: the problem occurred again while exiting an OS2 window session. The information they recorded and sent to me was: access violation at 1a0c44f3 OS2CHAR.DLL 0001:000044f3 Don't know if this happens in a full screen session. So far I believe it's only in an OS2 window. Dan Lee If anyone can recreate this problem, reopen the defect and let me know how you did it. Eric Farr 982-XXXX. ═══ 2.131. 60126 BETA: unwanted command reference icon in information folderduplicate ═══ name 60126 reference abstract BETA: unwanted command reference icon in information folderduplicate state canceled severity 3 answer invinf symptom vis duplicate defects: none. duplicate features: none. notes: If you EXCLUDE the online command reference during installation you still get the command reference icon in the information folder. I tried to reproduce this with 6.481, but the problem did not occur. To test, I installed 6.481, formatting the partition. At the main panel I entered documentation and deselected command reference, exited documentation and pressed "install". After rebooting I opened the information folder and found that "command reference" was not there, as was expected. In order to investigate this further we need more specific information, i.e. - what other features were selected/deselected, install with format vs installing over a previous version of OS/2, etc. ═══ 2.132. 60145 BETA: Wildcard searches using * fail ═══ name 60145 reference abstract BETA: Wildcard searches using * fail duplicate defect 55880 state canceled severity 3 answer duplicate symptom vis duplicate defects: none. duplicate features: none. notes: In View utility wildcard searches fail to exucute properly. For example a search of work* would fail to find working, worker workman, or works. This is a duplicate of defect 55880. The irony is that this fix was ready for 6.479 but the powers that be chose not to include it in the build. It is in 6.481. ═══ 2.133. 60150 BETA: Unwanted change of session type ═══ name 60150 reference abstract BETA: Unwanted change of session type duplicate state canceled severity 3 answer invsug symptom vis duplicate defects: none. duplicate features: none. notes: After a dos application was migrated. The session type was changed from Dos Window to Dos Fullscreen. Afterwards the name of a batch program was substituted for the program name. The session type was automatically converted back to DOS window. This seems to be the way the system always responds to a change in the program name. This is a good idea. We will consider it for the next post 2.1 for the next post 2.1 ═══ 2.134. 60163 BETA: Ver command still returns 2.00 ═══ name 60163 reference abstract BETA: Ver command still returns 2.00 duplicate state canceled severity 3 answer invinf symptom comp duplicate defects: none. duplicate features: none. notes: I *know* this seems like a nit, but several developers have called to tell us that all of the ways they have to determine the OS2 version still return 2.00 instead of 2.1. These include the dosversion functions in some C compilers, and the int21 function 30h. Dan Lee I really must insist that you open the defects required. MISC does not get this done. INT21 has a component, open one there. Versions are the H files I think, check with T.Lay. He is already working on this problem. Check with him for the correct component name. Also, please VERIFY/ACCEPT/CANCEL ═══ 2.135. 60167 BETA: EPM SYS 3175 access violation loading file ═══ name 60167 reference abstract BETA: EPM SYS 3175 access violation loading file duplicate state canceled severity 3 answer valunr symptom thipe duplicate defects: none. duplicate features: none. notes: BETA: I was trying to recreate a user reported problem and ran into this one. I made a copy of swapper.dat by using a DOS booted system. Then I boot 6.481 and start an OS/2 window session. I enter- epm swapper.zzz --where swapper.zzz is copy of swapper.dat. I get a SYS 3175 access violation for ETKE551.DLL. The E editor complains about the line length (maybe a clue) This problem is no longer reproducable. I suspect it was caused by a shortage of memory or selectors/address space. I'm already looking into those. (I supsect we have an innumerable number of these bugs.) When/If the problem occurs again, please record the CSLIM and IP. Also check if an entry was put in EPM.LOG before the trap. This would be in addition to the normal things like seeing if it is reproducable. After doing that, check if the problem is reproducable. This will help if the CSLIM/IP and EPM.LOG don't provide enough clues. I suspect I'll be updating the 32bit (and perhaps the 16 bit) version of EPM to dump the registers (and other useful data) to EPM.LOG upon trapping. In the meantime, I'm returning this defect as unreproducable. Please cancel it. ═══ 2.136. 60169 BETA: NULL embedded characters popup has truncated word ═══ name 60169 reference abstract BETA: NULL embedded characters popup has truncated word duplicate defect 62105 state canceled severity 3 answer duplicate symptom vis duplicate defects: none. duplicate features: none. notes: BETA: While trying to reproduce a problem I ran into this one. I tried to bring up the e editor along with the cmdref.inf file. I get a popup about null characters embedded in the file. The last word on the popup is severely truncated. I can only see some dots for the word "loaded". I believe the dots I see are the top of the letters "l,d,d". This is a duplicate of 57276 which was TRSd because it the fix requires an MRI change. It was caused by the change in fonts in R205 (it can be dupliated there also). TRSed defect reopened for Fit and Finish This is already fixed under an APAR - defect 62105 ═══ 2.137. 60198 BETA: DM_PRINTOBJECT DM_DISCARDOBJECT API's ═══ name 60198 reference abstract BETA: DM_PRINTOBJECT DM_DISCARDOBJECT API's duplicate state canceled severity 3 answer invtcp symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: DM_PRINTOBJECT and DM_DISCARDOBJECT API's are being passed invalid pointers for Dragitem and Draginfo respectively. A customer sent me a small executable and its source to demonstrate the problem. The exe opens a window and shows icons for files. If one runs this under GA and drags one of the icons to the shredder or a printer the file is removed or printed. If one runs the same executable under 6.479c or 6.481 and drags an item to the shredder or the printer the app is immediately closed rather than the object being shredded or printed. I have the code on a diskette and have not had time to acutually look, but the customer has debugged a bit and says that apparently invalid pointers are being passed to the Dragitem, Draginfo parameters. Please contact me a 982-XXXX for the exe,code. The customer's test program is freeing the draginfo structure on return from DrgDrag. At this time, the customer application is not guaranteed to have received the DM_PRINTOBJECT or DM_DISCARDOBJECT message. It works on GA because of a bug in GA that syncronized these messages incorrectly. (They are not specified as being syncronous, and in fact, we had a customer critical situation of them being syncronous.) James Taylor 982-XXXX ═══ 2.138. 60209 BETA: Can't disable Print Manager in WinOS2 ═══ name 60209 reference abstract BETA: Can't disable Print Manager in WinOS2 duplicate state canceled severity 3 answer chkmt symptom comp duplicate defects: none. duplicate features: none. notes: In GA it is possible to go into the Control Panel and Disable the WinOS2 print manager from the control panel. However under OS/2 Beta 2.1 the user in unable to disable the Print Manager from the control panel. I have talked with Joe Hlavaty about this problem prior to opening up this defect. This is a defect. But it is minor and does not prevent a WIN-OS/2 user from printing. The only effect is that the print manager icon continues to show up on the WIN-OS/2 desktop (as usual) even though it has been disabled. Therefore, this defect has been trs'd to be fixed a future release of OS/2. I discussed this problem and the temporary solution with Mike Jasinski and he agreed. -- Ayo Anise (2-6794) Defect needs to be marked "TRS" so that it will automatically show up against the next release. This problem will be fixed in the next release of OS/2. -- Ayo Anise ═══ 2.139. 60217 BETA: AmiPro 3.0 under OS/2 2.1 ═══ name 60217 reference abstract BETA: AmiPro 3.0 under OS/2 2.1 duplicate state canceled severity 3 answer invunr symptom use duplicate defects: none. duplicate features: none. notes: Another problem with Ami-Pro under Win-OS/2. If I go into help, help dows not go into the foreground, but starts up in the background. I have to shrink my AmiPro window and click on the help window to see it. Strangely, Excel 4.0 under Win-OS/2 does not have this problem. A BETA TESTER If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX. Defect 60217 is a duplicate of 59990. This is a duplicate of 5990 which was closed as unreproducible. This problem is reproducible. If you need assistance, please contact Sheila McDonald-Rich on 3-XXXX. Thanks The correct duplicate defect number is 58541. The same problem occurs with Lotus 123 and their help. Problem still exists in 492 driver. If any questions, please contact Sheila McDonald-Rich on 3-XXXX. Thanks I was show this on driver 6.492 using Amipro 3.0. Note that we could not reproduce it before because we were using Amipro 2.0 and could not find 3.0. This problem was fixed by a change to User and is now working correctly in driver 6.494. Mark Rogalski ═══ 2.140. 60220 BETA: AmiPro 3.0 will not go away ═══ name 60220 reference abstract BETA: AmiPro 3.0 will not go away duplicate defect 59989 state canceled severity 3 answer duplicate symptom use duplicate defects: none. duplicate features: none. notes: I started AmiPro 3.0 in a winos/2 windowed session. I started the AmiPro file management facility which started another window as it should. I then went back into AmiPro proper without ending Ami file management, finished my work and ended AmiPro. The File management window was still visible and the AmiPro Icon was shadowed showing it was still active. If I click anywhere on the file management window, I activated whatever was behind it on the desktop. If I clicked on the AmiPro icon, nothing happened. I tried to shutdown and got the "are you really sure" window and said yes. I got the prompt informing me that AmiPro was still active and I said OK to kill it. The system would not shut down. Further shut- downs did not realize Ami-Pro as active, at least I did not receive a prompt. I had to re-boot. Upon reboot, AmiPro restarted as it had been active at shutdown. A BETA TESTER If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX. Defect 60220 is a duplicate of 59989. ═══ 2.141. 60222 BETA: ATM Inactive ═══ name 60222 reference abstract BETA: ATM Inactive duplicate defect 60647 state canceled severity 3 answer duplicate symptom use duplicate defects: none. duplicate features: none. notes: Success! Broke the code! The testing of ATM was done on a system where OS/2 2.1 Beta was ADDED to a previously-installed OS/2 2.0 GA system. Now, installing ATM on a system with OS/2 2.1 "raw," everything is OK: o The version number appears on the ATM screen instead of "ATM Inactive" o ATMFONTS.QLC is built correctly in \PSFONTS o The ATM icon is NOT "stopped" when you start WIN-OS/2 Fullscreen All the developers have to do now is fix the install over a prior 2.0 system. Could it be due to the "OS2GDI.EXE" vs. "GDI.EXE" problem??? A BETA TESTER If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX. Ayo, please get this one cleaned up. Could you also see about taking over ownership of this component? Thanks. This is an install problem. When Borg is installed on GA (without formatting) ATM.INI does not get updated to include QLCDir=. This is a new addition to ATM.INI. I have discussed this problem with Pete Rodriquez. -- Ayo Anise This is a duplicate of 60647 which says the os2gdi.exe line is not written to system.in. ═══ 2.142. 60225 BETA: CDD.SYS driver trap 000d ═══ name 60225 reference abstract BETA: CDD.SYS driver trap 000d duplicate state canceled severity 3 answer invinf symptom thipe duplicate defects: none. duplicate features: none. notes: BETA: CDD.SYS is a public driver. Origin unknown. It allows one to be asked during boot time when the driver loads if one would like to to a command prompt rather than boot to workplace shell. It apparantly pauses the boot on loading for the user to hit . If the user does not hit on GA loading of WPS continues. On 6.479c and 6.481 if the user does not hit when loading continues it traps. I have the driver here if you need a copy. Brian Duquette 982-XXXX What was the location of the trap, trap screen info is very helpfully Also, the KERNEL component is a place holder for CMVC not a true OS/2 component. Please do not write defects against the KERNEL component, since files can NOT be delta'ed under it. Screen Team Approved ═══ 2.143. 60240 BETA: Internal error booting with debug kernel ═══ name 60240 reference abstract BETA: Internal error booting with debug kernel duplicate state canceled severity 3 answer invunr symptom thipe duplicate defects: none. duplicate features: none. notes: BETA: This problem is being submitted for Mike Turner (2-XXXX). Please contact Mike to recreate it. Please transfer this to the correct component if the current one is wrong ---------------------------------------------------------------------- Subject: Internal Error booting with Debug Kernel I am trying to install the debug kernel on my OS/2 2.1 system. However, during the boot process, after the Image Adapter device driver is loaded (I see the login messages), but before the HPFS386 (from LS 30 Advanced) is loaded, I get the following trap (sorry for the break in the copy/paste, but I wanted to stay within 72 columns - I also included some internal diagnostic messages from the debug kernel that were spit out during the load - there were lots of them...): pa=01000000 sz=00000000 va=00000000 sel=0000 fl=0001 of=00000000 ow=0000 pa=01000000 sz=03000000 va=00000000 sel=0000 fl=0002 of=00000000 ow=0000 pa=04000000 sz=00000000 va=00000000 sel=0000 fl=4000 of=00000000 ow=0000 SFence OK a20 pDHt rPIC j syi System Debugger 03/16/89 Н80386┘ Symbols linked (os2krnl) The system detected an internal processing error at location ##0170:fff55354 - 000f:00014354. line 2256, in file 4019 @(#)pgalloc.c 6.12 92/03/16 048600b4 - 01/12 19:23:15.25 Internal revision 6.479, 92/12/09 The system is stopped. Record the location number of the error and contact your service representative. eax=000005a0 ebx=fefa6c0c ecx=0000000c edx=00000000 eip=fff55354 esp=00005d46 ebp=00005d52 iopl=3 -- -- cs=0170 ss=0030 ds=0168 es=0168 fs=0000 gs=0000 cr2 0170:fff55354 f60618 test byte ptr Нesi┘, esi=ff1c0380 edi=0000006b -- nv up ei pl zr na pe nc =fe1a3000 cr3=001f3000 18 ds:ff1c0380=01 This is a PS/2 model 95, 64MB memory, IAA+6091, LS30, + CM/2 Beta, OS/2 2.1 (obviously), and I am trying to install the 2.1 toolkit debug kernel. My other bootable partition has OS/2GA+SP, and the SP debug kernel installed and ran O.K. there. The installed software is the same (IAA+6091, LS30, CM/2 Beta). Michael R. Turner I spoke with Mike Turner, and he told me that he has not been able to reproduce it. ═══ 2.144. *60246 BETA: CD-ROM not in drive crashes system ═══ name 60246 reference abstract BETA: CD-ROM not in drive crashes system duplicate state closed severity 3 answer asnm symptom vis duplicate defects: none. duplicate features: none. notes: Installed and ran Grolier Multi Media encyclopaedia perfectly in a dos full screen session. Then removed the CD-ROM from the drive. Entered command to run program from dos prompt. As usual the dos session returned with a SYS003 Error - Device Not Ready. Selected the option to END PROGRAM. Screen returned to opening for groliers encyclopaedia, which was completely garbled. Couldn't get task list up. Had to reboot. This program works fine in driver 6.490. I have made several fixes regarding CD-Roms since driver 6.479b, when this defect was opened. Once the program terminates it returns to the DOS full screen command line. The "garbled" screen is the result of an odd video mode being used. Pressing Ctrl-esc or alt-esc or entering "EXIT" work just fine. If the garbled screen is a problem, then this defect should be re-opened against video. I've informed Bill Bodin of this problem. He tells me that a video fix is going into driver 6.491 that might solve the garbled effect. If 491 does not solve it, then please speak with Bill about re-opening this defect. George Fulk returned original defect saying that in the current build the system no longer hangs but there is a video problem that still needs to be fixed before GA release. George Fulk after initial investigation returned the defect and asked that it be reassigned to the video component. Screen Team Approved This is a potential dup. of 59982 in which Susan will verify. There are two problems here: 1) The mode is not saved/restored because we generally don't handle bios mode 13 with chain bit reset properly in our transfer routines (vvga). 2) When the pop-up (CD-ROM not present) occurs, the mode is misidentified as ENHANCED and the wrong copy routine is called for that reason, which causes a trap. (vsvga). This has been fixed in 6.494. The Int15 AH=4 call scans through the ROM looking for ABIOS extensions. When it touches the XGA ROM a page fault occurs and VXGA freezes the session since it is not running in a DOS fullscreen. The XGA ROM contains the XGA coprocessor memory mapped registers and VXGA only allows the foreground fullscreen DOS app to access these registers since we do not virtualize the XGA coprocessor. Please reopen this defect. My answer and remarks are for a different defect. Sorry for the confusion. I have a fix for both the trap and save/restore of the mode. ═══ 2.145. *60267 BETA:Partial redraw problem with APM Power object ═══ name 60267 reference abstract BETA:Partial redraw problem with APM Power object duplicate state closed severity 3 answer dsai symptom vis duplicate defects: none. duplicate features: none. notes: BETA:This problem is submitted for Ken Inoue. Mindy is aware of the problem and according to Ken has reproduced it. ------------------------------------------------------------------- ----- OS2BETA FORUM appended at 04:24:49 on 93/01/13 GMT (by JL17356 at YMTVM1) Subject: Partial redraw problem with APM Power object There is apparently an off-by-one bug when the APM Power object processes WM_PAINT for partial window update. Sample scenario follows: (sorry for a longish append, developer explicitly asked for details. You needn't read further unless you have an APM enabled notebook PC) 1. Bring up the Power object, full status. 2. Open a windowed OS/2 command prompt. 3. Move the mouse pointer to the OS/2 window menu bar, click and hold the left mouse button, and move the window to partially obscure the Power object, then release the button. The windows may look e.g. as follows: .-------------------------------. |x| Power | |-------------------------------| |Battery life: Power source | .-------------------------------------------------. |x| OS/2 Window | |-------------------------------------------------| | | 4. Again move the pointer to the OS/2 window menu bar, click and hold the left button and move the OS/2 window elsewhere so that the Power object is completely exposed, then release the button. 5. Upon completing this scenario I see the following: .-------------------------------. |x| Power | |-------------------------------| |Battery life: Power source | | ---- | | / \ | | |\ /|: | | | ---- |: | | | :.-. |: (etc.) | | | :| | |: | | \ '-'/ : | | ---- | '-------------------------------' where ':' indicate black lines one dot wide, apparently left over from the OS/2 window (mostly black background). The color of the leftover garbage (':') matches that of the correspon- ding area in the obscuring window. In the above example, the upper right part of the garbage is light cyan (selected menubar's color). This doesn't occur if the Power object was completely obscured by the other window (thereby causing a complete redraw after the latter's removal). I would suspect problems with the partial update routine in WM_PAINT. I recall seeing something similar on beta version of Mahjongg v2.0 from Yorktown (uh..) so it could be the video driver (640x480x256 VGA256C). A side problem: When I tried to revert to 640x480x16 to see if this was video specific, DSPINSTL failed while trying to update some .INI file. This may not be so noticeable on monochrome LCD systems but is glaringly obvious on color LCD systems. Mindy told me they managed to recreate this problem & to report it here to get a defect writeup. The config is Thinkpad 700C with 4M memory, 120M harddisk, December 2.1 beta with 640x480x256 VGA256C 512K (system limit) installed via DSPINSTL APM support enabled at install. Hope this helps. Ken Inoue ═══ 2.146. *60274 BETA: serial device supp is default for selective install ═══ name 60274 reference abstract BETA: serial device supp is default for selective install duplicate state closed severity 3 answer chkm symptom use duplicate defects: none. duplicate features: none. notes: BETA: If one chooses a selective install, installing serial device support is selected by default. Then when one goes to install say video support through selective install, the Device=x:\os2\com.sys line is overwritten and any parameters ( n,xxxxx,i ) one might have chosen earlier is removed. This is a very bad idea especially when you consider a novice user may have no idea about this method of enableing com support. Add check to see if com.sys or vcom.sys lines already, if they do exist then install will not rewrite those lines. If the com.sys or vcom.sys lines are missing then install will update the config.sys. ═══ 2.147. *60275 BETA: sys crash when running clipboard with win/os2 image fduplicate ═══ name 60275 reference abstract BETA: sys crash when running clipboard with win/os2 image fduplicate state closed severity 3 answer lfm symptom vis duplicate defects: none. duplicate features: none. notes: Launched full screen win/os2 session. Ran windows version of Describe word processor. Drew a circle and cut to the windows clipboard. Opened the windows clipboard and viewed the graphic in three formats, bitmap, dib bitmap, and picture. Then clicked on OS/2 desktop. From the OS/2 desktop ran PM clipboard viewer. Machine traps out and reboot is necessary. Isn't a build break, isn't a system test blockage, by definition, it is SEV 3, as this is a defect process, not an APAR process. James Taylor 982-XXXX From the scenario that Marc described, this sounds like it is a dup of defect 59548...They are attempting to view a picture from a Windows** App which equates to a metafile in PM. The dupped defect # occurred when a user clipped a metafile and then opened the clipboard viewer, the system would hang. This was fixed in driver 6.481. Please attempt to reproduce on the current driver. Roberta 3-XXXX Debugged appears to be a PMCLIP problem ... Glenn has seen the trap ... Roberta 3-XXXX Compressed bitmaps where not being handled. No consideration was made to handle compressed bitmaps. ═══ 2.148. *60298 BETA: Help inoperative on closing window of DOS or OS2 sessduplicate ═══ name 60298 reference abstract BETA: Help inoperative on closing window of DOS or OS2 sessduplicate state verify severity 3 answer asnm symptom fnwad duplicate defects: duplicate features: none. notes: BETA: User has reported and I have verified that if you close a DOS or OS/2 window session, the help button does nothing on the popup that tells you there may be an active program in the session. Please transfer this to the appropriate component if I have indicated the wrong one. Thanks. The RES # is 6001 for this workplace shell help. This is one of many Help push buttons that was not enabled for GA. Tatchi, I noticed you returned 60070 as invsug. You may not want to fix it now but it is a valid problem, not just a sug. That message box has a help button on it. If you're not going to put up help, at least take away the button. --Mindy Pollack Enable Help for closing a session. ═══ 2.149. 60301 BETA: Sort function not working ═══ name 60301 reference abstract BETA: Sort function not working duplicate state canceled severity 3 answer invwad symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: This one's easy to reproduce. Just open up productivity. Open an applet. Say Calculator. Play around a bit. Then Close it. Select SORT by last access time. Things seem to move but PMFAX stays as the first object. None of the SORT by's seem to work. Sort by only works on objects that support that field in a column of the detail view. If you open productivity in the details view, you will see that FAX/PM, being a folder, and thus on the file system, is the only object that supports these columns. Abstract objects do not maintain this information. James Taylor 982-XXXX ═══ 2.150. *60306 BETA: SYS0206 RIPLINST on Disk 10 ═══ name 60306 reference abstract BETA: SYS0206 RIPLINST on Disk 10 duplicate state closed severity 3 answer regr symptom use duplicate defects: none. duplicate features: none. notes: SYS0206: The file name or extension is too long: C:\ibmlan\rpl\os2.20a\os2\mdos\winos2\systemnwpopup.exe Neat Name, but it aint gonna work on a FAT file system. - happened to NWPOPUP.EXE and PCSA.DRV A Beta Tester If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX. The destination field in the layout did not have a ending \, so newfit.exe interpreted it as a filename and not a directory. ═══ 2.151. 60351 BETA: AMAZE Farside Calender failing in beta ═══ name 60351 reference abstract BETA: AMAZE Farside Calender failing in beta duplicate defect 59741 state canceled severity 3 answer duplicate symptom fnwad duplicate defects: none. duplicate features: none. notes: Start program. In file menu -select NEW. In file menu - select SAVE AS ( to save the event file ). Enter a filename. Select OK. Message box appears stateing: 'Event file (filename) is marked as read-only and cannot be modified. This DOES NOT happen running in Windows 3.1 under DOS. In the Winos2 kernel, OS2K386.EXE, we open the file 1st time with read/write and share Deny_Write mode. This means only one file can open and read/write to a file. A second call to open read/write to a same file will be deny. This is fixed by allow Deny_None open mode attribute. This means any body now can open a file with Read/Write request. Joe Luu 26843 Duplicate defect report of 59741. ═══ 2.152. 60352 BETA: Street Atlas looses button bar when run seamless ═══ name 60352 reference abstract BETA: Street Atlas looses button bar when run seamless duplicate state canceled severity 3 answer valunr symptom vis duplicate defects: none. duplicate features: none. notes: Delorme Mapping's Street atlas runs properly in full screen Win/OS2 session. When run as a seamless windowed app on the OS/2 desktop button bar in the upper right of screen is missing. The screen area was painted white instead. I believe this is a duplicate of 58655 (dealing with z-order positioning of windows). If actual testing proves otherwise, then reject the answer. ═══ 2.153. 60357 BETA: Issuing a CTRL-C in a full screen disables Kboard induplicate defect 59591 ═══ name 60357 reference abstract BETA: Issuing a CTRL-C in a full screen disables Kboard induplicate defect 59591 state canceled severity 3 answer duplicate symptom thipe duplicate defects: none. duplicate features: none. notes: Borg Beta: 6C.479 When issuing a ctrl-c, the screen looses keyboard focus. Steps to recreate. 1. Bring up Full Screen. 2. Do a Ctrl-C 3. Type "Type config.sys | more" and press enter. 4. When you see "---more---" a message is received that you have lost kbd focus. 5. If I bring up file command, I can not use my keyboard, the keyboard is inactive. Any questions call me. Bob Gonzalez 982-XXXX This defect is a duplicate of defect 59591. ═══ 2.154. 60375 BETA: Running XGA 640X480X256, Get NO Bitmap Load in stall duplicate ═══ name 60375 reference abstract BETA: Running XGA 640X480X256, Get NO Bitmap Load in stall duplicate state canceled severity 3 answer valunr symptom fnwad duplicate defects: none. duplicate features: none. notes: I am running a Mod 95 with XGA, and after desktop opens the loading of a image to the desktop background stalls. I CNTRL-ESC to get a Task-list but get a END session thread. I choose Enter button to cancel session and it hangs. a script from debug can be made. Change to Wpsh , error sending to bkgrnd... This problem is caused by a problem on the build forum that says that all drivers before 6.485 will hang on a read > 64k. On the background page we do just that and when we do the DosClose, we hang. So wait for that driver and verify that it is unreproduceable. --Mindy Pollack Verified fixed in 6.486..Canceled Changed phase found to be " beta ". ═══ 2.155. 60383 BETA: Desktop icons do not maintain location after reboot. ═══ name 60383 reference abstract BETA: Desktop icons do not maintain location after reboot. duplicate defect 60047 state canceled severity 3 answer duplicate symptom fnwad duplicate defects: none. duplicate features: none. notes: Using non-grid format, scramble icons on screen. It returns to the install state and opening windows shows all icons half in window w/scroll bar. I have tried both sizes of icons, same result. I tried a flowed which was fine as well as non flowed. This looks like a build regression. This does not belong in ICON component... David A Kerr This one is fixed in 6.486. Also, when opening defects, sev 1 is build breaks, sev 2 is system test case blocks, sev 3 is important for borg and sev 4 is for suggestions or nits. -Sheila 2-XXXX Cancel duplication 60047. Changed phase found to be " beta ". ═══ 2.156. 60384 BETA: GPFAULT in OS2USER.EXE at 0001:17d6 ═══ name 60384 reference abstract BETA: GPFAULT in OS2USER.EXE at 0001:17d6 duplicate state closed severity 3 answer chkm symptom thipe duplicate defects: none. duplicate features: none. notes: BETA: ALT-TAB in winos2 fullscreen using enhanced mode produces .... Winfile caused a General Protection Fault in module OS2USER.EXE at 0001.17d6 to reproduce: open winos2 session in enhanced mode. open some windows. press Alt-F1. This occours in 6.479 as well as 6.481. This does not occour in Win 3.1 standard mode. Windows code assumed Enhanced Mode means VxDs are being used. Joe H. x3-XXXX ═══ 2.157. *60394 BETA: Wrong Bit Map Data Is Being Stored in Metafile ═══ name 60394 reference abstract BETA: Wrong Bit Map Data Is Being Stored in Metafile duplicate state closed severity 3 answer dsam symptom dc duplicate defects: none. duplicate features: none. notes: Test program supplied by a developer, BITMTST3. During debug of problems associated with this program is was discovered that BitMap data stored in the metafile is corrupted. Joe Celi has a copy of the demo program. Joe Celi has requested that this defect be opened. If we are converting an internal bitmap into external 4bpp format, we should use the logical color table within the DC if it has exactly 16 entries. Joe Celi ═══ 2.158. 60399 BETA: Install fails on Multimedia machines ═══ name 60399 reference abstract BETA: Install fails on Multimedia machines duplicate state canceled severity 3 answer valunr symptom comp duplicate defects: none. duplicate features: none. notes: On some Model 57 (77?) Multi-media machines, 2.1 installation returns the error "System could not find ABIOS.SYS". If this file is copied manually, the system will continue to return messages about files it could not find. If you keep copying files manually you can eventually get to where the system tries to reboot off its hard disk. At this point you get the "can't find country.sys" message. Dan Lee Screen Team Approved Dan, since we have not ever seen this problem with the Beta or after, I will have to assume that this problem was as we discussed - the user inserted the wrong diskette. I have since changed the check to be slightly more stringent on which diskette we will accept so that this (hopefully) will not happen again. ═══ 2.159. 60400 BETA: MSMONEY Has Only One Digit Right Of Decimal Point ═══ name 60400 reference abstract BETA: MSMONEY Has Only One Digit Right Of Decimal Point duplicate state closed severity 3 answer chki symptom dc duplicate defects: none. duplicate features: none. notes: Microsoft Windows 3.1 provides support for 2 digits right of the decimal point in International Currency as a default. WINOS2 only provides 1 digit to the right of the decimal. WINOS2 International Currency Default needs to be changed to 2 digits to the right of the decimal point. Lisa Garcia is aware that I am opening this defect. ═══ 2.160. 60429 BETA:Daily Planner has Save the Changes problem ═══ name 60429 reference abstract BETA:Daily Planner has Save the Changes problem duplicate state canceled severity 3 answer chkit symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: User has submitted this problem and I have been able to reproduce it. ---------------------------------------------------------------------- ----- OS2BETA FORUM appended at 20:34:14 on 93/01/13 GMT (by USA1 at ALMVMA) -- Subject: Bug in Daily Planner... If you are running daily planner and make a change, and then move to the next day, the program says changes have been made and asks if you want to save them. This is good. However, if you say no, it will let you go to the next day, but when you try to change days again, it asks you again if you want to save the changes, even if no new ones were made. I guess the program is still registering the changes that were not saved from before and asking if they should be saved. Anyhow, it should not do this. If the person changes days and doesn't save the changes they should be lost after the first prompting. Mark Carlson USA1@ALMADEN.IBM.COM The answer really is consider for next release. Function still works. Prog. asking a question that it should not ask at that parti cular point in time. ═══ 2.161. 60460 BETA: Failure to start 2nd inst of OS/2 Window from dstp poduplicate ═══ name 60460 reference abstract BETA: Failure to start 2nd inst of OS/2 Window from dstp poduplicate state canceled severity 3 answer invwad symptom vis duplicate defects: none. duplicate features: none. notes: Problem starting the second instance of an OS/2 window from the OS/2 Desktop pop up menu. Settings for the OS/2 window program were changed to allow for new instances of the program. This works fine for clicking on the desktop icon. Add the OS/2 Window program to appear on the desktop popup menu. It never opens up a second instance, but rather takes you back to first instance. This behavior has been evident since 2.0. the purpose of the menu page is to run programs as a view of the object who owns the menu not the program object that appears on the menu. So, if you want multiple occurences of the program to appear, you have to set it on the object not the program. So, in the case of the desktop popup menu, set multiple views on the desktop not on the OS/2 window. Mindy Pollack 2-XXXX ═══ 2.162. 60480 BETA: Seamless Winos2 icons creep up desktop ═══ name 60480 reference abstract BETA: Seamless Winos2 icons creep up desktop duplicate defect 57818 state canceled severity 3 answer duplicate symptom vis duplicate defects: none. duplicate features: none. notes: BETA: When a seamless winos2 session is minimized to the desktop, if one opens and minimizes it continuously, the icon will "creep" up the desktop. To reproduce: pull a template to the desktop, put progman.exe or any other winos2 app in and select winos2 window. Also select minimize window to the desktop from window page. Then open progman or whatever, minimize it, reopen it from the windows list rather than the icon, minimize it, do this continuously and you will notice the icon working it's way up. This does not happen for other types of windows minimized to the desktop. minimized windows to the desktop are owned by pmwin. --Mindy ═══ 2.163. *60511 BETA: lost cursor in dos fullscreen with > 28 lines ═══ name 60511 reference abstract BETA: lost cursor in dos fullscreen with > 28 lines duplicate state closed severity 3 answer lfm symptom vis duplicate defects: none. duplicate features: none. notes: BETA: If one issues mode co80,50( anything > 28 lines ) in a DOS fullscreen using Trident 8900c video adapter. The cursor is lost. This happens in all resolutions, VGA, SVGA but not with other video adapters. Functionally, all commands work well and characters typed do appear, just the cursor is missing. The video for DOS Emulation is handled through video VDDs. I am re-routine this from DEM to VVID. This problem is partly due to the scaling algorithm for the cursor in procedure BvsSetCurType of file src\bvfcalls\bvfcursr.asm, and partly due to the Trident cards. It should be all right to have the cursor thickness represented by equal starting and ending positions. Most SVGA cards show one scan line as the cursor in the DOS FullScreen, but Trident does not show the cursor at all. This is a hardware problem of this card. One way to solve this problem is to use 2 scan lines to represent the cursor, under the condition that the application which uses this routine means to have a visible cursor. This change won't affect the case where the cursor is supposed to be invisible. ═══ 2.164. *60518 BETA: selective install does not allow path ═══ name 60518 reference abstract BETA: selective install does not allow path duplicate state closed severity 3 answer intm symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: If one installs from the complete set of diskettes and then decides they would like to selectively install from a different source( CD-ROM. LAN, etc. ) one cannot select a path for the diskette images but is forced to insert the diskettes. This is not consistant as if one installed from CD-ROM or LAN and chooses a selective install on is allowed to choose the path\medium of installation. This needs to be consistant. While not as much of a concern for 2.0, we now ship on several media and this could be a concern. ═══ 2.165. 60552 BETA: Tree view changes font of Icon View ═══ name 60552 reference abstract BETA: Tree view changes font of Icon View duplicate defect 59304 state canceled severity 4 answer duplicate symptom vis duplicate defects: none. duplicate features: none. notes: Merely doing a "TREE VIEW" of any folder - say the desktop - will change the font of Icon view. Use button 2 on the desktop Use MB2 select 'OPEN' followed by Tree view. The desktop fonts immediately change to I think System Proportional. this is the same problem as 59304 and was fixed in borg driver 6.496 make that driver 6.486 DUplicate of 59304 - fixed in 6.486 ═══ 2.166. 60563 BETA: SYS0047 from pipe to more ═══ name 60563 reference abstract BETA: SYS0047 from pipe to more duplicate state canceled severity 3 answer symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: If one uses cntrl + break at a press any key to continue prompt from /p and then attempts to issue a pipe to more the message: SYS0047: The keyboard focus is unavailable SYS1805: The process trued to write to a non-existant pipe To reproduce: Open OS/2 fullscreen type dir /p at the press any key to continue prompt press ctrl+break type type config.sys | more ═══ 2.167. 60583 BETA: TRAP 0008 when shutting down the system - ═══ name 60583 reference abstract BETA: TRAP 0008 when shutting down the system - duplicate state returned severity 3 answer invinf symptom thipe duplicate defects: none. duplicate features: none. notes: BETA: Customer gets a TRAP 0008 when doing a Shutdown. There are several scenarios under which this problem has happened. However, the TRAP 0008 does not happend consistently under these scenarios. One of them consists of creating a folder, shredding it and then shutting down. Another scenario is sending a note through Golden Commpass, then shutting down. Here is the trap information provided by the user: TRAP 0008 ERRCD=000 ERACC=**** ERLIM=******** EAX=0000000D EBX=7d03700e ECX=00000140 EDX=00000440 ESI=fff20158 EDI=7d037000 EBP=012c5ecc FLG=00012202 CS:EIP=0f00:00003068 CSACC=009b CSLIM=0000ad6f DS=0158 DSACC=c093 DSLIM=ffffffff CR0=8001001b ES=0158 ESACC=c093 ESLIM=ffffffff CR2=00135480 FS=0000 FSACC=**** FSLIM=******** GS=0000 GSACC=**** GSLIM=******** location ##0160:fff60967 - 000d:9967 048600b4 Internal revision 6.479, 92/12/09 We need more information to be able to do anything with this. What applications were running on the system at the time? What had actually happened in the shutdown before the trap occurred (e.g. which popups had been displayed and what applications had already been terminated)? Please try to gather as much information as possible. Donald Hobern 2-XXXX Screen Team Approved ═══ 2.168. 60585 BETA: System hangs when running seamless app - Codewright ═══ name 60585 reference abstract BETA: System hangs when running seamless app - Codewright duplicate defect 60607 state canceled severity 3 answer duplicate symptom vis duplicate defects: none. duplicate features: none. notes: Was running two OS/2 applets, Chess and Pulse, along with the Win/OS2 clock and Codewright seamlessly. Codewright is a Windows 3.1 programmers text editor. I played a bit of chess, changed the forground and backround color of pulse, changed the clock to analog and back to digital, changed fonts. Opened a second editing window in Codewright (which is a MDI app). Switched between Codewright's editing windows added text, cut and pasted, did undo's then system hung. This scenario was created in the OEM lab on a Compaq Deskpro 386/33 on the latest 486 driver. These two defects describe the same problems. The descriptions in these defects simply describe scenarios. It is not clear to me that we can assume the winos2 has a defect in a scenario that involves the wpshell, multiple os/2 apps and a winos2 app. The submitting team needs to debug these problems to isolate the failing component. The problem described in 60585 is described in 60607. ═══ 2.169. *60601 Beta: clipboard does not display Metafiles ═══ name 60601 reference abstract Beta: clipboard does not display Metafiles duplicate state closed severity 3 answer lfi symptom fnwad duplicate defects: none. duplicate features: none. notes: Created by CMVC to address defect 59709. BETA: When one copies a metafile to the clipboard, the title Bar says Metafile yet render will not display the the Metafile in the clipboard. Document# s61g-XXXX-00 OS/2 2.1 Using the Operating System on page 18-31 states: "When a metafile is copied to the clipboard, it is converted into a bitmap, and then the image is enlarged. Only the center of the image is in the clipboard viewer. This will be fixed in a later release." The metafile is in the clipboard because it can be pasted to another picture viewer, but this document seems to imply that some image should be in the clipbaord viewer when there is none. please contact Lisa Garcia for a detailed description of why I rejected this defect. Apparantly she has identified problems that are not clearly visable at first glance. CMVC:Fixed incorrect AFLD2 field. Had to set the presentation space to PU_PELS before playing the metafile. And change the calculation for scrolling the metafile. Roberta 3-XXXX ═══ 2.170. 60602 BETA: Background and Lockup Directory ═══ name 60602 reference abstract BETA: Background and Lockup Directory duplicate state canceled severity 4 answer invsug symptom use duplicate defects: none. duplicate features: none. notes: I'd like the ability to concatenate another directory (or more) to \OS2\BITMAP, especially for the purpose of putting backgrounds on other than the boot disk. This is all part of my "general aversion" to placing "user data" on the "operating system's" disk. As it is, if you install a new operating system with the "format" option, you must remember to copy your bitmap backgrounds to the newly installed system. Or, if you have multiple operating systems on your disks, you must have multiple copies of the bitmaps. If I were given a choice between a "SET BITMAP=..." entry in CONFIG.SYS or the ability to specify a directory in the "Background" page of a notebook, I'd probably go with the former (but really don't have a strong opinion). NOTE: Even though \OS2\BITMAP appears in "SET DPATH=...", other directories in the DPATH are NOT searched for backgrounds. ~~~~~ If DPATH were searched, no problem... A Beta Tester If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX. let's sug this one and see what happens in the next release. Mindy I typed in an incorrect defect number (it should be 60620) when I accepted the defect. And thus, I have to reassign it to you. I apologize for any inconvenience it may cause. Shelia, I returning this defect as a suggestion. However, we will fix this in the next release. Suggestion was forwarded to planning ═══ 2.171. 60603 BETA: Double-click or Open in Start Here? ═══ name 60603 reference abstract BETA: Double-click or Open in Start Here? duplicate state canceled severity 3 answer chkmt symptom use duplicate defects: none. duplicate features: none. notes: The phrase "double-click" has been replaced "open" in the USING.INF file. Out of curiosity, I went into the "OS/2 Tutorial" and the "Start Here" objects to see how consistently the terms were used in OS/2 2.1 (how long has it been since YOU, dear reader, actually looked at either of them?). The "OS/2 Tutorial" starts off with "...select a topic on the right by double-clicking on it..." where one "topic on the right" is the picture of the mouse. Selecting that item, very early the text reads "To select (open) an object, place the pointer on the object and double-click mouse button 1...." After that, the "OS/2 Tutorial" seems to be consistent in the use of the term "open." Now go to the "Start Here" icon. I quickly found about ten references (using "Search") to "double-click." In fact, the references all seemed to read "...double-click mouse button 1 (the left button)...." I *think* that the intent of "Start Here" should be to emphasize the term "open" at all of these references. Use of "button 1" and/or "the left button" is to be avoided, if I understand the "OS/2 Tutorial" and OS/2 2.1 Using the Operating System" (USING.INF). Subject: Double-click or Open in Start Here? Ref: Append at 22:11:53 on 93/01/12 GMT (by DEV4163 at HONE84) In the referenced append I commented about how the phrase "double-click" has been replaced "open" in the USING.INF file. Out of curiosity, I went into the "OS/2 Tutorial" and the "Start Here" objects to see how consistently the terms were used in OS/2 2.1 (how long has it been since YOU, dear reader, actually looked at either of them?). The "OS/2 Tutorial" starts off with "...select a topic on the right by double-clicking on it..." where one "topic on the right" is the picture of the mouse. Selecting that item, very early the text reads "To select (open) an object, place the pointer on the object and double-click mouse button 1...." After that, the "OS/2 Tutorial" seems to be consistent in the use of the term "open." Now go to the "Start Here" icon. I quickly found about ten references (using "Search") to "double-click." In fact, the references all seemed to read "...double-click mouse button 1 (the left button)...." I *think* that the intent of "Start Here" should be to emphasize the term "open" at all of these references. Use of "button 1" and/or "the left button" is to be avoided, if I understand the "OS/2 Tutorial" and OS/2 2.1 A Beta Tester If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX. TRS next release. Per Marion Lindsey. TRS'd ═══ 2.172. 60604 BETA: Environment Variables -- Proposal ═══ name 60604 reference abstract BETA: Environment Variables -- Proposal duplicate state canceled severity 4 answer invwad symptom use duplicate defects: none. duplicate features: none. notes: This is a proposal ,please consider: I spend a lot of time examining issues regarding OS/2 flexibility. The issue I wish to address today is that of "Environment Variables". I have carefully examined the use of environment variables in OS/2. I have some proposals which should improve the capabilities of the OS/2 system, while maintaining, if not improving, existing system performance levels. A significant problem in OS/2 is that the PATH, LIBPATH, and DPATH environment variables can become excessively long. At presnt, my LIBPATH exceeds 400 characters. This is a real PAIN to manage (EPM can't edit it), and negatively impacts system performance. There has to be a better way. I propose to implement four categories of environment variables, update the SET command, and provide enhanced program access to the environment variables. The four categories of environment variables are: System - Established in CONFIG.SYS Application - Established via Appl ID Notebook Settings (New Feature of Program Object) Logon - Established via User ID NoteBook Settings (New Feature of User Profile Management) Instance - Established via Command Line Request (Existing Feature, SET Command) In this setting (please forgive the pun, it was unintentional ), the SET command can display any category of environment variable, but can not update the System, Application, or Logon categories. Each category is to be maintained via its respective interface. If an environment variable is specified in more than one category, then the values will be concatenated in a system-defined order. We may wish to consider whether to permit the concatenation order to be determined by the user. This could be a system configuration option. The SET command with no operands is currently used to display the values of all defined environment variables. I suggest that four new flags be added to limit the display to a specific category of environment variable. The flags associated with each environment variable category are: System - /S Application - /A Logon - /L Instance - /I If no flag is specified, then the concatenated value will be displayed. New functions should be provided to scan the environment variable area. I suggest the following function definitions be used as a starting point: typedef LHANDLE HSCANENV ; HSCANENV DosBeginScanEnv( ULONG ulEnvCat ) ; APIRET DosScanEnvNext ( HSCANENV hScan , PSZ pszEnvVarName , PSZ pszResultPointer ) ; APIRET DosEndScanEnv ( HSCANENV hScan ) ; These function calls will be used in place of the existing DosScanEnv function call. The specifics of the function call interface are: DosBeginScanEnv is invoked, specifying a category of environment variables to be retrieved. If successful, it will return a handle to be used on subsequent function calls. The categories which may be specified in the DosBeginScanEnv function call are to be selected from the following list: ENVCAT_NONE ENVCAT_SYSTEM ENVCAT_APPL ENVCAT_LOGON ENVCAT_INSTANCE DosScanEnvNext is invoked specifying the handle returned by DosBeginScanEnv, and pointers to buffers where the environment variable name and the environment variable value will be placed. Each subsequent invocation will locate the next environment variable within the category specified in the DosBeginScanEnv function call. When DosScanEnvNext indicates that no more environment variables can be located, the caller should invoke DosEndScanEnv to terminate the scan process cleanly. How does this proposal help ? The System PATH can be limited to those directories necessary to system startup. Search time is reduced, and overall system performance is improved. This consideration applies equally well to the System LIBPATH and System DPATH. WPS objects can then expand these variables to meet application- specific requirements. Overall system environment space is reduced, resulting in more available physical memory. Swapping is reduced, and once again, overall system performance is improved. Are these performance gains significant ? No, probably not. But they should more than make up for the effort expended in providing this additional functionality to the users. Implementation of this proposal will request changes to the User Profile Management component (UPM), which is included in Communications Manager, Database Manager, and LAN Server. All terminology in this proposal is subject to change, in order to meet IBM product standards. A Beta Tester If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX. Sorry, we can not do product enhancements as defects. Returning as working as designed. Please talk to product planning and get a feature raised for product enhancements. This customer has some valid points, but these are suggestions for product enhancement - these need to go to planning. Paul. WAD-Suggestion was forwarded to planning ═══ 2.173. 60605 BETA: Glossary and Master Index -- No SWAP Topic ═══ name 60605 reference abstract BETA: Glossary and Master Index -- No SWAP Topic duplicate state canceled severity 3 answer asnip symptom use duplicate defects: none. duplicate features: none. notes: Given my recent interest in the "swap" topic, I searched both the on-line Glossary and the on-line Master Index for "swap" and "*swap*" and got no hits. I suggest that *some* mention of this topic be included someplace other than the on-line Command Reference. (By the way, there is no tab in either document for "page" or "paging.") A Beta Tester If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX. Information will be provided in the Master Help Index in next release. We presently document in the Using the Operating System book and the Installation Guide. To be addressed in the next release ═══ 2.174. 60607 BETA: System hangs when exiting last seamless windows 3.1 aduplicate ═══ name 60607 reference abstract BETA: System hangs when exiting last seamless windows 3.1 aduplicate state canceled severity 3 answer invinf symptom vis duplicate defects: duplicate features: none. notes: After monitoring internal as well as ISV/CUST OS2BETA forums it became evident that a pattern existed regarding system hangs when running OS/2 apps and seamless windows 3.1 apps together. At times, upon exiting the last windows 3.1 app the system would hang. Forum appends were made by individuals with diverse video cards, drivers, processors, and apps. This problem was reproduced on Lisa Garcia's test machine using Pulse and Chess as the OS/2 apps and Premia Corps Codewright, a windows 3.1 MDI app. This bug was in the 12/92 beta as well as the most current driver 6.486. Joe, Ivan and I looked at this and found the following: The first thread of "Pulse" owns "fsrsUser" and is waiting for "_FSRDriverSem". "_FSRDriverSem" is owned by TID 1 of the Workplace Shell (PID 4) and is waiting for "ksemMFT". "ksemMFT" is owned by the first thread TID 1 of "pmshell" (PID 2) who is waiting for "fsrsUser", hence the deadlock. Marc Schneiderman has the scenerio to recreate this hang. It is too long to write down here so I told him that you would give him a call when you get this. His number is 3-XXXX. Thanks, Glenn Puchtel (3-XXXX) This is not a WINOS2 problem. The hang is due to a deadlock between PM and the display driver. There is no information to indicate that this problem is other than an XGA problem since it has not been debuged on another system. You need to reproduce this problem again. At that time I would be happy to assist the appropriate display or PM developer in debugging this deadlock. Glenn Puchtel (3-XXXX) ═══ 2.175. 60610 BETA: Seamless app causes PM debugger to vanish mysteriouslduplicate ═══ name 60610 reference abstract BETA: Seamless app causes PM debugger to vanish mysteriouslduplicate state working severity 3 answer chki symptom vis duplicate defects: none. duplicate features: none. notes: When using the C/Set 2 presentation manager debugger with Codewright running as a seamless app the debugger mysteriously vanishes. The C/Set 2 PM debugger is CSD #30 (Dec. 92). It works flawlessly with multiple open OS/2 apps, and an open OS/2 window for compiling and linking. The specific app that was being debugged when this behavior occured was the hello program from the toolkit. OS/2 Level 3.0 Change Team is working on this, contact Joe Wieloch 3-XXXX The PM Debugger IPMD.EXE posts a destroy window to the debugger source window causing the source window to be destroyed. Since the application is destroying the window, this is an application error and should be assigned to the PM Debugger development team. They should be able to provide insight into why the application is posting a destroy message. -- OS2 Change Team (Joe Wieloch 8-982-6261 / 3-6261) Screen Team Approved Roger Early (8-545-XXXX), a PM Debugger developer, called me today. The debugger is doing a DosDebug call and receiving an invalid notification code (303) in the debug buffer. He said that this was identified in the beta of PM Debugger 2.0 and that Arthur Roberts has created a fix. I talked to Arthur who delta'd code against defect 61066 to fix this problem, the fix will be in the Borg 6.492 build. PM Debugger ver 1.0 with the CSD will need to be tested on the 6.492 build. -- OS2 Change Team (Joe Wieloch 3-6261) Problem of running C/Set2 PM debugger along with a seamless windows app still persists in the 6.492 driver. In addition the Borland C++ Turbo Debugger for OS/2 also crashes when tracing thru an app that makes a call to create a new window. The test scenerio originally included the PM debugger of C/Set2 but after more extensive testing it seems that the beta version of Borland's C++ Turbo Debugger also exhibits abnormal behavior when stepping thru a sample program which is making a call to create a window. This behavior only manifests itself while a seamless app is running in the backround. The Borland PM Turbo Debugger exhibits abnormal behavior when stepping thru source code that creates a window when a seamless app is in the backround. Rich Wesson of Borland International was contacted 2-9-93 and made aware of this situation. His number is 408-438-XXXX ext 606. Spoke to James Boch, QA Mgr for the Borland Turbo debugger product. He said they were targeting OS/2 2.0 for their upcomming release, and were not actively testing under the 2.1 BETA. He indicated Borland's priorities did not include addressing this issue to resolution at the present time. Joe Wieloch spoke to Arthur Roberts, who is responsible for DosDebug. Arthur explained DosDebug calls only reflect the state of the system. It is up Lexington, KY and Borland Int'l to handle special cases. Currently the IBM PM debugger developers are working on a new release and are aware of the problem with running seamless apps concurently on the OS/2 desktop. Per information stated by Marc Schneiderman. Please note the following phone numbers: PM Debbuger: Patrick Gerstle 8-545-XXXX Roger Early 8-545-XXXX OS2 DosDebug: Arthur Roberts 8-982-XXXX OS2 Chng Team: Marc Schneiderman 8-982-XXXX Joe Wieloch 8-982-XXXX Spoke to Roger Early who felt the problem with the C/Set 2 debugger, when running a seamless app, was not application related. Since the same problem exists in the Borland C++ Turbo Debugger at exactly the same point, when tracing through a call to WinCreateStdWindow I am reopening this defect. Both the IBM PM debugger and Borland Turbo debugger step through calls to WinCreateStdWindow fine under 2.0 and Service Pak. DosDebug is notifying the PM Debugger that system resources are unavailable. To determine the real problem, it is necessary to get more information about what calls are made to DosDebug and what DosDebug is checking/returning. Marc S. also tested on the 4.96 build with a 1/20/93 version of IPMD.EXE, in this case the system locked up, another indication that the problem might not be in winos2. -- Joe Wieloch 3-6261 OS2 Change Team . This is not an error condition. Rodger Early in lexington can explain what changed in his debugger to cause this to appear. The system resource which is being held is a semiphore two people can not control the same system resource at the same time. If you are in PM or any other area which requires screen changes or updates this is a possible message. If you need all of the details ask Rodger for the Apar number. Please route directly to Glen Puchtel. He has worked on debugging this problem with Dave Medina and has a fix. It seems that we (PM) are holding the user semaphore (fsrsUSER) at the time we need to do a DosLoadModule on pmviop.dll. This causes problems for debuggers so a method needs to be identified to load the module without having the user semaphore. So far the only safe place to do this is when the thread creates its message queue (ie WinCreateMessageQueue), at this time we will load "pmviop". This will change the timing somewhat, hopefully this timing change will not cause any regressions. ═══ 2.176. 60611 BETA: Sound from MVDM not shut off after task switch from pduplicate ═══ name 60611 reference abstract BETA: Sound from MVDM not shut off after task switch from pduplicate state canceled severity 3 answer invwad symptom fnwad duplicate defects: none. duplicate features: none. notes: Running SB, M-Audio.. ALT-ESC to change session to desktop still has music from app playing in background. This is a major bug and needs to be delt with ASAP. Known since GA 2.0. Changed phase found to be " beta ". OS/2 has no way of knowing what the sound card is doing, which is the one driving the speakers. The DOS app, in the other hand, is the one driving the sound card. There must be a way to have OS/2 sense the port activity and shut it down if the data is deemed to be of an address which was used by a killed process. I think this is a major for many customers, I for one. Cancel due to future enhancement projects. ═══ 2.177. 60621 BETA:SETUP.TXT file needs to be removed from WinOS2 ═══ name 60621 reference abstract BETA:SETUP.TXT file needs to be removed from WinOS2 duplicate defect 60649 state canceled severity 3 answer duplicate symptom vis duplicate defects: none. duplicate features: none. notes: The text file SETUP.TXT is in WinOS2 on the 6.486 build. I talked to Ayo Anise and he said that this file needs to be removed from the WINOS2 layout. This defect is being opened at Ayo's request. I have fixed this under defect #60649. Don Wakley 2-XXXX. ═══ 2.178. 60628 BETA: Win Timeout garbles after OS2 Fullscreen ═══ name 60628 reference abstract BETA: Win Timeout garbles after OS2 Fullscreen duplicate state closed severity 3 answer lfm symptom comp duplicate defects: none. duplicate features: none. notes: In a Seamless WinOS2 session, select Control Panel, Desktop, Timeout, and set the timeout value to 1 minute. Alt-Esc out to the Desktop and go into an OS2 full screen session, and wait there for 1 minute (without moving mouse or touching the keyboard). After the timeout has had time to kick in, Alt-Esc or Ctl-Esc back to the desktop. One of two things have been reported to happen: either the Ctl-Esc, Alt-Esc, or ...-Tab combinations don't let you out of OS2 full screen (you must type "exit"), or else they do let you back to the desktop but the WinOS2 seamless session is completely garbled. Dan Lee 2-XXXX Control Panel/Desktop/ScreenSaver Name Flying Windows Delay 1 Minute I have my 57sx with XGA, (8514 monitor), switch to OS/2 Full Screen and wait pass 1 minute. Ctrl-Esc back to WPS, notice that seamless Winos2 section all garble. I suspect that the video device driver or screen save need to know that it need to repain. Ctrl-Esc is being eaten up by the OS/2 and never being detect by Windows. Joe Luu DO NOT OPEN DEFECTS UNDER THE WIN31 RELEASE!!!!!!!!!!!!!! Level 3 change team will investigate. Screen Team Approved This defect is being assigned to Ivan Lomelli. Intercept Int2f to determine if in OS/2 or DOS full screen. When in full screen, prevent "isuseridle" function from starting screen saver. ═══ 2.179. 60630 BETA: SBP default board settings initialize, but breakup .Wduplicate defect 59648 ═══ name 60630 reference abstract BETA: SBP default board settings initialize, but breakup .Wduplicate defect 59648 state canceled severity 3 answer duplicate symptom fnwad duplicate defects: none. duplicate features: none. notes: Running SBP under default settings which is the ONLY way to use the Speech option in KQ6 initializes fine. When a .WAV is played it is prone to stuttering. If the SBP is set to I/O=240 IRQ=5 DMA=1 the load program in KQ6 for .WAV's cannot initialize. This is a major problem which needs to be in the next release. Fixed in defect 59648 against VDMA Duplicate on Defect #59648 on VDMA. fixed in 6.486. ═══ 2.180. *60634 BETA: Diskette's fail when scsi device presence check disabduplicate ═══ name 60634 reference abstract BETA: Diskette's fail when scsi device presence check disabduplicate state closed severity 3 answer chki symptom thipe duplicate defects: none. duplicate features: none. notes: BETA: With a PS/2 configured to have a scsi device and presence checking disabled for that device, if the device is powered off, boot from diskette will fail on diskette #1 attempting to do I/O to the device. As per Daily Mgr's.Status Mtg., 01/21/93, B.Odle directed the project office to route to Ken Rowland The EBDA was used to determine what SCSI devices are in the system. Since there are entries in the EBDA for powered off devices (only on IBM systems) the ADD was creating an entry in its internal tables for the device. When a DMD requested the list of attached devices the ADD returned this powered off device. The table entry, however, was never initialized to the correct state of the device. The DASD manager therefore thought that a DASD device was present. This caused the system to hang. Marking the device as defective fixes the problem. ═══ 2.181. *60647 SVT BETA : os2gdi.exe line not written to system.ini ═══ name 60647 reference abstract SVT BETA : os2gdi.exe line not written to system.ini duplicate state closed severity 3 answer dsai symptom appf duplicate defects: duplicate features: none. notes: os2gdi.exe line not written to system.ini when Borg is installed over a GA + Service Pak machine. Lisa 2-XXXX ═══ 2.182. *60649 BETA:Online Help Incorrect For Change System Settings ═══ name 60649 reference abstract BETA:Online Help Incorrect For Change System Settings duplicate state closed severity 3 answer inti symptom vis duplicate defects: duplicate features: none. notes: The WinOS2 Online Help for Change System Settings talks about setting up mouse, displays, LAN, etc. when the only option allowed is to set up the LAN connection. Ayo Anise is aware that I am opening this defect and said that this section will be rewritten. Reference to setting not support by WIN-OS/2 3.1 to be removed from SETUP.HLP -- Ayo Anise ═══ 2.183. 60653 BETA: Can't format IBM optical disk media ═══ name 60653 reference abstract BETA: Can't format IBM optical disk media duplicate state canceled severity 1 answer invunr symptom vis duplicate defects: none. duplicate features: none. notes: Can't format 3.5" IBM optical drive media. Used format j: /1. The computer responded ... The new type of file system is FAT. Insert a new diskette in drive j: and press enter when ready. Pressed enter at this point and SYS1267 appears. The system is unable to determine the optical disc type. Format appears to have been modified (around 5/92) to require a newer level of OPTICAL.SYS than is currently in he field. These modifications need to be adjusted so that FORMAT will work with the current level of OPTICAL.SYS as well as the new level since we do not ship the OPTICAL driver as part of OS/2. See phase1in.c after call to RWOpticalMediumSense() and this routine in api.c This problem occurred because a downlevel optical device driver was being used. A diskette sent from a customer (Ken Singer, Shell Oil Co.) was used to determine this. Marc Schneiderman has been given a copy of the correct device driver with instructions regarding installation of the same. ═══ 2.184. *60659 BETA: This is similar to ═══ name 60659 reference abstract BETA: This is similar to duplicate state closed severity 2 answer asni symptom thipe duplicate defects: none. duplicate features: none. notes: This is a re-opening of a problem that was fixed in defect 25316. The`symptoms are identical (according to the IBM developer in Rome (DCAF). I have a copy of the app and will get whatever detail is needed. I suspect that the fix got "lost" in r206. BETA: This is a re-opening of a problem that was fixed in defect 25316. Symptoms are identical (according to the IBM developer in Rome (DCAF). I have a copy of the app and will get whatever detail is needed. I suspect that the fix got "lost" in r206. This was written against mmm but the fix was applied to ifsm in r202, transferring to ifsm. Routing per discussion with Paul D. G. Simco Well the fix is in a different place, os it would seem to be a different problem. Paul. CMVC:Fixed incorrect AFLD2 field. CMVC: cloned to new defect 61370. ═══ 2.185. 60671 beta: tree command allows no path to be entered ═══ name 60671 reference abstract beta: tree command allows no path to be entered duplicate state canceled severity 3 answer intmt symptom fnwad duplicate defects: none. duplicate features: none. notes: Do tree /? Use the tree command to display the directory structure of a drive or path. tree c:\os2 Incorrect parameter is used is displayed Command reference has NO path option. According to B.Odle, this component belongs to you. Thanks. This goes agianst the tree utility. GA documentation stating that paths are supported for Tree was incorrect, this functionality was never in place. Will be worked for next release. Reopened for reevaluation Although it would be possible to add the functionality into the code, the documentation, including hardcopy pubs would have to be changed to complete the fix, and so this will be post- poned to the next release. cancelling for now... should be in the next release ═══ 2.186. 60677 BETA: 'Title Clash' in System Settings does not work ═══ name 60677 reference abstract BETA: 'Title Clash' in System Settings does not work duplicate state canceled severity 3 answer invsug symptom fnwad duplicate defects: none. duplicate features: none. notes: t prompted if he/she is using objects from the Template Folder. If he/she is using objects not in the Template Folder, then the user is prompted for a desired title. e.g. Drag an item from the Template Folder onto the desktop. If the operation is repeated, then the system automatically renames the second object. the shell is designed so that name clash dialog only appears for a copy or move. Or for a create another because the user can change the name. It was never implemented for templates. We should probably consider designing it to be more consistent in the future but it is not the current design. We cannot change design for this build. This defect will rollover in r207 --Mindy Pollack ═══ 2.187. 60694 BETA: COMMAND.COM SYS3006 error ═══ name 60694 reference abstract BETA: COMMAND.COM SYS3006 error duplicate state working severity 3 answer docm symptom use duplicate defects: none. duplicate features: none. notes: I added an item to the Desktop menu to start a DOS prompt. The program name is COMMAND.COM. When I en and path info from the name when I save it. Anyway, when I select DOS from my menu the DOS sessions starts but I always get a SYS3006 error message (the drive or path specified with the COMMAND.COM is not valid). I'm also getting the same error when I start a Windows 3.1 application with an AUTOEXEC.BAT file specified. A Beta Tester If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX. This problem was reported 2 weeks ago and fixed then. I have driver 6.486 installed, and the problems addressed in this defect seem to me to be fixed. Screen Team Approved Subject: COMMAND.COM SYS3006 error The problem is still reproducible on the 492 driver. If you need further assistance, please contact Sheila McDonald-Rich on 3-XXXX. This problem is caused by WorkPlace Shell. Unforuntately, this is also working as designed. WorkPlace Shell associates objects, not files. When the "DOS Prompt" is started WorkPlace sees no intructions in the Program Arguments of the COMMAND.COM object so it uses the default action; which is to pass the folder name as a parameter. The net result is COMMAND.COM is being passed "C:\DESKTOP" as a parameter. COMMAND.COM interrupts the passed parameter as the location for the COMSPEC environment variable. COMMAND.COM issues the error message because it can't find COMMAND.COM in the C:\DESKTOP sub-directory. So how do you do this. WorkPlace Shell has an undocumented feature to pass no parameters. You need to open the Drives folder, open OS2, open MDOS, and lastly open the settings to COMMAND.COM. On the 2nd line (arguments), place a percent sign (%). This is the undocumented way WorkPlace knows not to pass a parameter. Personal opinion: This should not be undocumented. I'd go so far as to say a defect could be opened against ID to document this. This concerns a development problem. This defect should be reassigned to DOS development or WPSH development. (I discovered this same problem). Vince Spadea ID 2-XXXX Problem appears to be in WPSH. If any questions, please contact Sheila McDonald-Rich on 3-XXXX. Thanks Vince, Per my conversation with Peter Magid, I am rerouting this defect back to you. Please contact him as he will provide you with the technical information that needs to be added to the online help. Thanks, ═══ 2.188. 60695 BETA: OS/2 Logo in the Scramble Game is incorrect ═══ name 60695 reference abstract BETA: OS/2 Logo in the Scramble Game is incorrect duplicate state open severity 4 answer symptom use duplicate defects: none. duplicate features: none. notes: Subject: Bogus OS/2 Three-Ring Logo Another example of an OS/2 Logo wannabe is in the Scramble game. Seledt "Game" then "Open..." then "OS/2 Logo" to see what I mean (among other details, there is an extra ring in this logo). A Beta Tester If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX. Screen Team Disapproved Change Team: Delegated work item! ═══ 2.189. 60696 BETA: Behavior of Icons Iin VIEW Window ═══ name 60696 reference abstract BETA: Behavior of Icons Iin VIEW Window duplicate state canceled severity 3 answer invwad symptom use duplicate defects: none. duplicate features: none. notes: If you VIEW an .INF file, CMDREF.INF for instance, and "minimize" a few of your viewed pages (say, the table of contents), you can move those icons all over the VIEW window. If you drag the icon over the left, right, or top edges of the VIEW window, it pops back "reasonably" and the text is completely visible. However, if you drag it to the extreme bottom of the VIEW window, the icon pops back and overlays the text accompanying the icon. Also... if you have an icon or two in the VIEW window, then resize the window in such a way as to "loose" some of the icons, I cannot see a way to make them all appear short of maximizing the window. If you shrink the size of that window vertically, icons get pushed off the bottom of the window. A Beta Tester If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX. The problems you describe are all due to the MDI design of the help window. This is correct behavior for an MDI application. We recognize that the MDI model has problems and we intend to address these in some future release by eliminating the MDI model but currently, this is "working as designed". WAD-to be addressed in a future release ═══ 2.190. 60707 BETA: Cannot create boot floppies if using 4DOS ═══ name 60707 reference abstract BETA: Cannot create boot floppies if using 4DOS duplicate state canceled severity 4 answer invnos symptom fnwad duplicate defects: none. duplicate features: none. notes: Part of the CD/ROM installation process is creating the boot floppies. If the program 4DOS is used instead of COMMAND.COM, the step that creates these diskettes will fail. The failure occurs at the point where you are asked to insert the blank diskette and press any key to continue. When you press the key, an error message immediately appears that says "Possible error creating the OS/2 2.1 installation diskette". The program 4DOS is marketed by JP Software. I contacted them and they investigated and believe the problem is in the MKINDSKF.EXE routine. They can be reached at 617-646-XXXX The contact's name is Tom Rawson. the program that creates diskettes from the borg CD was not developed by the Install group. This was a special program only for the borg beta. It will not be part of the shipped product. It was produced by a team led by Eva Cannova, pls contact her. ═══ 2.191. 60709 BETA: Model 70 dies on boot ═══ name 60709 reference abstract BETA: Model 70 dies on boot duplicate state canceled severity 3 answer invinf symptom vis duplicate defects: none. duplicate features: none. notes: Customer has Model 70 with scsi quantum Drivekit 245 as well as ESDI controller. Installed using diskettes with sysinst2.exe dated 12-26-92. The system hangs after booting to OS/2 logo is displayed. Steve Best has copy of exact error message. When the system boots, immediately after the OS/2 logo is displayed, the customer gets the following message: The system detected an internal processing error at location ##0160:fffba480 - 000d:00063480 1157, 7019 0386000d1 Please ask the customer for registers when this traps occurs. With the information available it can not be determined where in the kernel this trap is occurring. Steve Best 3-XXXX The trap occurred in file vdmm.c (7019) on line 1157. The code was testing for a bad return code (most likely 8) from VMAllocKHB. We should not IPE for an out of memory condition. The trap occurred in file vdmm.c (7019) on line 1157. The code was getting a bad return code (most likely 8) from VMAllocKHB. If the rc was 8, we should not IPE, since its only an out of memory condition. ═══ 2.192. *60710 Beta : INT_DURING_IO setting hangs system ═══ name 60710 reference abstract Beta : INT_DURING_IO setting hangs system duplicate state closed severity 3 answer dsai symptom thipe duplicate defects: none. duplicate features: none. notes: BETA: When the INT_DURING_IO setting is used for uninterupted sound with multimedia applications the system hangs. VPIC's semaphore was being posted at terminate time. The slave thread thought it had a valid request. Most likely, parts of memory were destroyed based on this bogus request. VPIC now validates the request before attempting. If the correct flag is not set, it exits the routine. ═══ 2.193. *60724 BETA: A Leading Space on Lock Pass Word Causes Problem ═══ name 60724 reference abstract BETA: A Leading Space on Lock Pass Word Causes Problem duplicate state closed severity 3 answer inti symptom fnwad duplicate defects: none. duplicate features: none. notes: I have tested the following and can reproduce the problem on the 6.486 build. Mike Jasinski Subject: Lockup Problem It appears a password which begins with a space screws up the help function on the locked screen. Given such a password, if the system is locked and if help is then selected and cancelled, there is no way to have the password accepted. Entry of the first space will return the help window. Only way out is to reboot, and that doesn't have very nice results since there was no orderly shutdown. do not write any defects against wpsl please. ═══ 2.194. 60741 BETA: Install confuses D & C drives. ═══ name 60741 reference abstract BETA: Install confuses D & C drives. duplicate state canceled severity 3 answer invinf symptom appf duplicate defects: none. duplicate features: none. notes: CorelDraw: Trying to install ADOBE type manager on D drive - install assumes it is on C drive. Problem on 2.1 Beta - CorelDraw for OS/2 & Windows. The defect does not have enough useful information to debug the problem. The submitter needs to add more information in order to debug this problem. Screen Team Approved I have CorelDraw 3.0 for Windows, and ATM Adobe doesn't come with it. If you are using OS/2 2.1, and installed the OS/2 operating system on default driver (which is C:), it automatically creates a C:\PSFONTS directory. And the Winos2 ATM version 2.5 default is set to C: also. All this information is kept under winos2\ATM.INI. Now the question unclear in the problem is if you already installed ATM Adobe and it resides on drive D:, and you install the operating system on driver C:, then the ATM is still default to C: as described in the previous paragraph. This is working as design. There is no entry during OS/2 2.1 installation that asks the user whether he has ATM adobe fonts installed or not. The only thing the user can do is to start ATM, and run ADD option. Now if you installed OS/2 2.1 Beta on the default driver D:, then \PSFONTS will be created on D:. Joe Luu ═══ 2.195. 60744 BETA:Install hangs on diskette 1 on AMBRA machine ═══ name 60744 reference abstract BETA:Install hangs on diskette 1 on AMBRA machine duplicate defect 58329 state canceled severity 3 answer duplicate symptom thipe duplicate defects: none. duplicate features: none. notes: BETA:Several users have reported the Install process hangs on Diskette 1 on an AMBRA machine. Craig Bennett thinks we didn't see problem here because our machines are downlevel. Paul Devriendt is also aware of problem. This problem has now been reproduced in the OEM Lab in Boca, and is the same as defect 58329. Paul. ═══ 2.196. *60746 BETA: starting windows app from startup.cmd corrupts desktoduplicate ═══ name 60746 reference abstract BETA: starting windows app from startup.cmd corrupts desktoduplicate state closed severity 3 answer asnm symptom vis duplicate defects: duplicate features: none. notes: Starting a windows app from startup.cmd corrupts the full screen Win/OS2 session. If the app was migrated then the OS/2 desktop is corrupted. This was tested on an IBM Valuepoint with Trident SVGA drivers in the performance test lab. this defect should not be a sev. 1. Sev. 1 is for build breaks, sev 2 is for system test case blocks and sev 3 is for needs to be fixed in BORG if the problem exists when you start the app via the startup.cmd, then this isn't a shell problem. --Mindy Pollack 2-XXXX OS2 Level Change Team currently working on this problem, contact OS/2 Level 3.0 Change Team is working on this, contact Joe Wieloch 3-XXXX. This problem also occurs when you try to start any WINOS2 app from the OS2 Window command prompt (i. e. start c:\os2\mdos\winos2\clock.exe). Investigating. On Trident, when starting WINOS2 minimized with VIDEO_SWITCH_NOTIFICATION set to OFF, winos2 comes up corrupted. This is not observed on other adapters. If VIDEO_SWITCH_NOTIFICATION is set to ON, an additional BIOS mode set is issued when VDM comes into foreground, so that the corruption is taken care off by the winos2 driver itself. When starting a WINOS2 app in the background (using start) and when going into the foreground, winos2 is corrupted and there is no additional mode set to take care of the corruption (in other words, it behaves as VIDEO_ SWITCH_NOTIFICATION set to OFF). Settings of that app's icon or WINOS2 settings have no impact on this behaviour. Desktop corruption on a very first switch from the WINOS2 FS to the desktop happens only if the app was started from an OS/2 FS and a switch to the desktop occured before the WINOS2 session was brought full screen. The desktop dirty flag flVVD is set by the VVID in the background when first BIOS mode set happened, but because we switched to the desktop first, the flag was reset. If the subsequent BIOS mode set existed, when the vdm session went full screen, the flag would have been set again. This is possibly a design flow in the usage of the flag on part of both desktop and vvid. Conclusion: If when starting WINOS2 minimized without the swith notification set can not be fixed by VVID, this defect is out of our realm. Will investigate why is that corruption happening in the first place. I will also investigate why is the flag reset on a swith from a non VDM session to the desktop. On Trident, when starting WINOS2 minimized with VIDEO_SWITCH_NOTIFICATION set to OFF, winos2 comes up corrupted. This is not observed on other adapters. If VIDEO_SWITCH_NOTIFICATION is set to ON, an additional BIOS mode set is issued when VDM comes into foreground, so that the corruption is taken care off by the winos2 driver itself. When starting a WINOS2 app in the background (using start) and when going into the foreground, winos2 is corrupted and there is no additional mode set to take care of the corruption (in other words, it behaves as VIDEO_ SWITCH_NOTIFICATION set to OFF). Settings of that app's icon or WINOS2 settings have no impact on this behaviour. Desktop corruption on a very first switch from the WINOS2 FS to the desktop happens only if the app was started from an OS/2 FS and a switch to the desktop occured before the WINOS2 session was brought full screen. The desktop dirty flag flVVD is set by the VVID in the background when first BIOS mode set happened, but because we switched to the desktop first, the flag was reset. If the subsequent BIOS mode set existed, when the vdm session went full screen, the flag would have been set again. This is possibly a design flow in the usage of the flag on part of both desktop and vvid. Conclusion: If when starting WINOS2 minimized without the swith notification set can not be fixed by VVID, this defect is out of our realm. Will investigate why is that corruption happening in the first place. I will also investigate why is the flag reset on a swith from a non VDM session to the desktop. Screen Team Approved Freeze enhanced mode session after a mode set in the bgnd. ═══ 2.197. 60754 BETA: performance problem - excessive time on DLL loads. ═══ name 60754 reference abstract BETA: performance problem - excessive time on DLL loads. duplicate defect 59295 state canceled severity 3 answer duplicate symptom prf duplicate defects: none. duplicate features: none. notes: This defect was written to insure that a private fix to 2.1 BETA sent to SAS Institute to improve 2.1 BETA run times on DLL loads is included in the next release. I am at a loss to understand why you wrote this defect. As you know, this was a known problem, that was already fixed before you found the problem. Please check the database for writing defects for known problems. Thank you. Paul. Screen Team Approved THe problem has been resolved ═══ 2.198. 60760 BETA: Driver for IBM Proprinter ═══ name 60760 reference abstract BETA: Driver for IBM Proprinter duplicate state canceled severity 4 answer valunr symptom use duplicate defects: none. duplicate features: none. notes: The entry for "4201 ProPrinter" appears "out of order" in the PRDESC.LIST. It follows the II and III entries. A Beta Tester If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX. Rerouted to install. Screen Team Approved This has been fixed for the 6.494 build. The list will be sorted correctly in that build. Proprinter will come before Proprinter II, which will come before Proprinter III, etc. I changed the build tool so that the colon is not considered as one of the characters to sort in the sort algorithm... ═══ 2.199. 60761 BETA: Errors in PRDESC.LIST ═══ name 60761 reference abstract BETA: Errors in PRDESC.LIST duplicate state canceled severity 4 answer invwad symptom use duplicate defects: none. duplicate features: none. notes: In the list of printers, PRDESC.LST, there are two entries for the 4019 and four entries for the 4029 that are in error. The entries that call for "LASERJET.DRV" refer to "Laserprinter" (in two places for each entry) instead of "LaserPrinter." Note the bicapitalization? "LaserPrinter" is not a trademark (according to the manuals that come with one) but whatever rights IBM might have to the term would be best protected by correcting those entries. Besides, all the other references use "LaserPrinter." FWIW, "LaserJet" is a trademark of HP. BiCapitalization -- the industry ran out of acronyms, or acronyms went out of style, so now two common words get pressed together. OfficeVision, MultiMedia, BookMaster (and BookMaster/2!); A Beta Tester If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX. Rerouted to install. Rerouting from INS to LaserJet. Devnames takes the name from the LaserJet driver itself (retruned from DevQueryDeviceNames). I don't imagine you can change the name now, because we already shipped with it and there would be problems on the network (drivers on the server having one name and drivers on the requestor having another name)... I'll let you and Mark decide this one... Thanks, Kim Rerouting from INS to LaserJet. Devnames takes the name from the LaserJet driver itself (retruned from DevQueryDeviceNames). I don't imagine you can change the name now, because we already shipped with it and there would be problems on the network (drivers on the server having one name and drivers on the requestor having another name)... I'll let you and Mark decide this one... Thanks, Kim Sorry about the double text... the first time I rerouted it, CMVC told me there was a problem and it didn't work... I checked an IBM LaserPrinter, and the P in Printer is upper. The names of IBM LaserPrinters do have tradmark errors in Laserjet driver. Unfortunitily, we can not change them. Otherwise, mix use of laserjet drivers with different versions in the lan envirnment will generate errors. ═══ 2.200. *60776 BETA: Child windows in AWS eat up 8 meg of shared address sduplicate ═══ name 60776 reference abstract BETA: Child windows in AWS eat up 8 meg of shared address sduplicate state closed severity 3 answer dsai symptom prf duplicate defects: none. duplicate features: none. notes: BETA: From Mark Cates and David Jeans at SAS institute. Creating child windows in a application worksspace uses up to 8Meg of shared memory for each child window created. SAS sent us sample software to illustrate this, which has been turned over to Rick Thompson to chase down. Screen Team Approved Changed from a single very larg segment to multiple segments in the segment store. Rick Thompson 982-XXXX ═══ 2.201. 60783 BETA:Help does not function when diskette drive b: noduplicate ═══ name 60783 reference abstract BETA:Help does not function when diskette drive b: noduplicate state canceled severity 4 answer intit symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: I have no diskette drive b on my system. If I open the drives folder and double click on drive b icon, I get popup about putting a diskette in the drive. This popup has a help button. If you push the button, nothing happens. I should see a help screen. The GA version put up a different error message, and the help button does function on that popup. Please transfer this to the correct component if I have picked the wrong one. This defect was written against the wrong component. I am transferring it to the message file. This is a prompt message. Prompt messages do not have helps. Working as designed. Anita 2-XXXX Spoke to Mindy Pollack in Development. This is a hard error and should be reassigned to SSMG. Anita 2-XXXX. This problem was inadvertantly returned to me. It should be re-opened. It is a regression from the 2.0 GA product. The user sees a different message than before which may be okay, but the popup has a help button and that button does nothing. In the GA product, the help button gives you a help screen. Bob, this problem is not a regression from GA. My developer talked this over with Paul Cheatham and we will TRS to the next release. Please talk to Paul before rejecting this answer. THANKS! I've cancelled it, but it is truly a regression from GA. In the same scenario, the GA product puts out a different message and the help button does give you a help screen. In this version, you get a different message and the help button is inoperative. ═══ 2.202. 60788 BETA: Machine Hangs at OS/2 logo after boot mgr screen ═══ name 60788 reference abstract BETA: Machine Hangs at OS/2 logo after boot mgr screen duplicate state canceled severity 3 answer symptom vis duplicate defects: none. duplicate features: none. notes: Machine hangs at OS/2 screen after selecting partition to boot from boot manager menu. If user hits reset button on PC (not CTRL-ALT-DEL) machine will boot fine from subsequent boot mgr screen selection. Configuration is as follows: Milex MAE486 33mhz system unit Milex SCSI DCE376 controller Fujitsu 500mb hard disk GSI model 11 floppy controller with 1.2 mb, 1.44mb and 2.88 drives attached Milex GLE911 video card with 1mb ram using vanilla VGA drivers. Future domain 850 SCSI with texel 3024 CD Rom PAS - 16 board nothing attached Serial card with COM1 and COM2 ports (Tiawanese $20 special) Parellel board - LPT1 Screen Team Approved ═══ 2.203. 60797 SVT REG: TRAP E during execution of Rexx ═══ name 60797 reference abstract SVT REG: TRAP E during execution of Rexx duplicate state closed severity 2 answer chki symptom thipe duplicate defects: duplicate features: none. notes: Customer Impact: The system is trapping Hardware: Model 95 Reproduce the problem: Install the driver as normal. Add the following to the config.sys: set lib=hellolib set tmp=hellotmp set include=helloinc Reboot the system. Execute a batch file called RXRGSTR1. A trap E is generated. You must power OFF the system to restart your system. For more information contact Bernice W. Graham This problem could not be reproduced in the lastest BORG build driver level 6.490 . Screen Team Approved The problem still exist on driver 6.494 and 6.496. Attached is a snapshot of the defect. Invalid Data is passed in the DS register. The trap is in REXX.DLL at sys_free within the REXX and PMREXX component. The following information is supplied: Testcase Name: RXRGSTR2 Testcase Name: REG316 ##ln 0170:fff47509 os2krnl:DOSHIGH32CODE:HaltInst + 1 0170:fff475fb postSchedNext - f1 Symbols linked (rexx) Symbols linked (rexxapi) Trap 13 (0DH) - General Protection Fault 0000 eax=162e35b0 ebx=000cf6e4 ecx=fffffffc edx=00000000 esi=00000000 edi=162e35b0 eip=16248955 esp=000cf6a0 ebp=000cf6bc iopl=2 rf -- -- nv up ei pl nz na po nc cs=005b ss=006b ds=0053 es=0053 fs=150b gs=0000 cr2=0015c000 cr3=001d1000 005b:16248955 8b11 mov edx,dword ptr [ecx] ds:fffffffc=invalid ##.ln 005b:16248948 rexx:CODE32:sys_free + d 005b:16248a3c sys_realloc - e7 ##.k 005b:161e6ea6 00000002 0017f8c4 162e35b0 00000002 H_rxfree + 3a 005b:161cbfbf 162e35b0 162e3a94 00000002 001b0f14 sys_external + 8b 005b:161d6b56 162e35b0 162e3a94 00000002 001b0f14 subroutine + 37a 005b:161d74c2 0017fb68 162e35b0 001821c2 001821c8 H_logical_to_int + 62a 005b:161d6479 162e35b0 0017fb68 00000000 00000001 expression + c5 005b:161d63c9 162e35b0 0017fb68 00000000 001821b6 expression + 15 005b:161e038e 162e35b0 00000005 0017fc48 00000000 interpret + da 005b:161de085 162e35b0 0017fc34 00000000 1a02a3a8 instruction + d71 005b:161ec163 162e35b0 00000001 00191ca1 162e35b0 user_routine + 1bf 005b:161d6a6f 162e35b0 00191bf9 0017fccc 00000000 subroutine + 293 005b:161dd920 0017fd90 162e35b0 0019177f 00191785 instruction + 60c 005b:161dfe4a 162e35b0 00000002 00191017 00000002 instruction + 2b36 005b:161dd870 162e35b0 0017fee8 00000012 00000003 instruction + 55c 005b:161ec163 162e35b0 00000001 00190a41 162e35b0 user_routine + 1bf 005b:161ebe87 162e35b0 00190a41 0017ffb8 00000001 rexxprogram + 3ab 005b:161c13da 162e35b0 162e3880 00000000 00000000 RexxMain + 42a 005b:161a0100 00000000 00000000 00040258 00000000 005b:000d09dc 00000000 00183000 00000000 0000003f 005b:161c0890 00000000 00000000 00040258 00000000 RexxStart32 + 1ec 005b:161a0041 00000000 00000000 00040258 00000000 rexxdllterm + 31 ************ NOTE: REG316 is trapping at the same location and address in REXX. Testcase Name: REG316 ## r Trap 13 (0DH) - General Protection Fault 0000 eax=162e35b0 ebx=0017f8c4 ecx=fffffffc edx=00000000 esi=00000000 edi=162e35b0 eip=161e8955 esp=0017f880 ebp=0017f89c iopl=2 rf -- -- nv up ei pl nz na pe nc cs=005b ss=006b ds=0053 es=0053 fs=150b gs=0000 cr2=001ec000 cr3=001d1000 005b:161e8955 8b11 mov edx,dword ptr [ecx] ds:fffffffc=invalid ##. .ln Invalid option ##ln 005b:161e8948 rexx:CODE32:sys_free + d 005b:161e8a3c sys_realloc - e7 ##.pb Slot Sta BlockID Name Type Addr Symbol 0001 blk fff13dc8 *ager 0002 blk fff79971 *tsd _tkTSDDaemon 0003 blk fff45be4 *ctxh _kmCTXHDaemon 0004 blk fff79e72 *kdb _tkKDBDaemon 0005 blk fff04b9c *lazyw _semLW 0006 blk fff12c94 *asyncr _AsyncReadSem 0008 blk fffe000d pmshell RamSem 0763:06d6 000a blk ffca0002 pmshell 000b blk fffd000b pmshell MuxWait 000c blk fffd000c pmshell MuxWait 000d blk 04000a90 pmshell DosSem 0400:0a90 CtrlNumLkQ 0007 blk fe90ffdc pmshell Sem32 8001 0019 vhevLazyWrite 0010 blk fea2d374 pmshell Sem32 8001 0001 SrvReq 0012 blk fffd0012 pmshell MuxWait 0013 blk fffe0007 pmshell RamSem d09f:0bd0 memory_pool + 127 0014 blk fffe0008 pmshell RamSem d09f:0bd8 memory_pool + 12f 0015 blk fffe0009 pmshell RamSem d0c7:0050 0016 blk fffd0016 pmshell MuxWait 0017 blk 04000a82 pmshell DosSem 0400:0a82 SMHotKeyQ 0018 blk fffd0018 pmshell MuxWait 0019 blk fffe000a pmshell RamSem 06c8:0b16 SEMLW 001a blk fffe000b pmshell RamSem 06c8:0c86 SEMRA 001b blk fffe000f pmshell RamSem 0763:06f2 Slot Sta BlockID Name Type Addr Symbol 0009 blk 04000aac harderr DosSem 0400:0aac HardErrPkt 000e blk 04000dc0 harderr DosSem 0400:0dc0 CtrlPrtScQ 000f blk 04000de4 harderr DosSem 0400:0de4 VideoPauseFlag 001c blk fffe0011 pmshell RamSem 0763:06f6 001d blk fecfbf58 pmshell Sem32 8001 0022 hevLazyWrite 001e blk fffe0010 pmshell RamSem 0763:06fa 001f blk fe9fa8f4 pmshell Sem32 0001 0005 hevSleeper 0020 blk fe9fa8c4 pmshell Sem32 0001 0007 vhevAger 0021 blk fffe0013 pmshell RamSem d2c7:0b4e SEMSPOOLERSCHEDULE 0022 blk 0c280012 pmshell 0023 blk 0c30000e pmshell 0024 blk fffe0014 pmshell RamSem d2c7:1f40 0026 blk 0c501000 pmshell 0027 blk fffe0015 pmshell RamSem d0c7:0088 0025 blk 05400678 cmd 0028 blk fffe001f cmd RamSem d0c7:00c0 002a blk fffe003f pmrexx RamSem 002f:00ea ## Problem first found in System Test. The Rexx.DLL is trapping due to incorrect checking of the length of the Rexx string being freed and in processing default strings. The problem has been corrected and a new Rexx.DLL is being integrated. ═══ 2.204. 60805 Orchid Prodesigner 2S goes out of sync after diskette 19 ═══ name 60805 reference abstract Orchid Prodesigner 2S goes out of sync after diskette 19 duplicate state canceled severity 3 answer invinf symptom vis duplicate defects: none. duplicate features: none. notes: During selective install, Orchid Prodesigner 2S display goes bananas after diskette 19. Messages say - Changing configuration......Checking Adapter setup...., then the display starts getting squiggly lines..,etc. When a reboot is performed, the system comes up in VGA mode. Customer has tried the DSPINSTL command but still fails. This defect should be updated with more information. If you could get in touch with your customer and ask the following questions: What are the video subsystem statements for VIO_VGA or VIO_SVGA and VDM video driver like? Are BVHSVGA and VSVGA statements present? Second, when SVGA WHO executed in DOS full screen, does it successfully identify the adapter as Tseng ET4000? If so, when running SVGA ON, is SVGADATA.PMI file generated properly, and does it list all expected resolutions (256 color and 16 color)? If so, after the file is generated and VGA desktop rebooted, is SVGADATA.ERR file present on \OS2 directory? If so, what does it say. Also, since the installation abruptly stopped, could it be that there was no room left to copy BVHSVGA and VSVGA onto the hard disk? And what exactly did you mean by DSPINSTL failed as well? In what fashion and at which point? Please, contact me at 3-XXXX if more information needed. Contacted customer and he indicated that he's too busy to send the information we need to continue on this, so we may as well cancel it. Screen Team Approved ═══ 2.205. *60810 Help button on display driver window doesn't work ═══ name 60810 reference abstract Help button on display driver window doesn't work duplicate state closed severity 3 answer asni symptom fnwad duplicate defects: none. duplicate features: none. notes: If one attempts to use DSPINSTL either by command line or through selective install, the help button on the window does not work. Help for display resolutions box was being handled improperly. Fixed by recreating help instance and letting the help manager process help messages. Screen Team Approved ═══ 2.206. *60814 BETA:Migrated config.sys shows FEAT in IFS= statement and tduplicate ═══ name 60814 reference abstract BETA:Migrated config.sys shows FEAT in IFS= statement and tduplicate state closed severity 3 answer chkm symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA:User Nelson Chen In Boulder has reported his migrated config.sys has FEAT inserted in some config.sys statements. Specifically IFS= shows FEAT: <\FDISK> and also SET BOOKSHELF= has FEAT:;<\TUTORIAL>. This leads to a trap e on boot. Craig Bennett has the config.sys and install logs and asked me to write a defect for the problem. He said he knows the fix. The problem was caused by two threads accessing the same global memory at the same time - one to log output and the other to update CONFIG.SYS. The log routines were debug statements that were interspersed with valid output, so these were removed. ═══ 2.207. 60816 BETA: Drive A icon notebook settings can not be opened ═══ name 60816 reference abstract BETA: Drive A icon notebook settings can not be opened duplicate state canceled severity 4 answer invwad symptom fnwad duplicate defects: none. duplicate features: none. notes: You can not open the settings notebook on the Drive A icon without having a diskette in the drive. This is a duplicate of cancelled defect 25502. You cannot open the settings view of an empty disk drive because the details page requires reading the drive. You get a hard error saying that the drive is empty. This is working as designed. --Sheila 2-8252 ═══ 2.208. 60822 BETA: XGA32 garbles 2.1 screen ═══ name 60822 reference abstract BETA: XGA32 garbles 2.1 screen duplicate state canceled severity 3 answer invibmhdwe symptom vis duplicate defects: none. duplicate features: none. notes: PS2 model 90, XGA on motherboard, XGA/2 *pre-production* card (as I spoke with Joe on the phone before reluctantly opening this) in slot two. Uses 640x480x65000 color mode, was okay on a 32-bit GRE build we sent him in June. Under 2.1, however, when he boots the system, he sees the OS2 logos okay but when the WPS comes up it's garbled as though it were very very badly out of sync. He does not have the XGASETUP.PRO anywhere under the XGA$DMQS subdirectory. (Meanwhile, back at the ranch...) I've asked the customer to try to find a production XGA/2 card to use, and let me know what happens. Dan Lee Screen Team Approved This problem was caused by a pre-production IBM XGA/2 card. Joe Celi ═══ 2.209. *60847 BETA: Writing to a Metafile Causes SVGA Driver To Hang ═══ name 60847 reference abstract BETA: Writing to a Metafile Causes SVGA Driver To Hang duplicate state closed severity 3 answer dsai symptom thipe duplicate defects: none. duplicate features: none. notes: Customer's PM application BITMTST5.EXE causes SVGA driver to hang when writing to a metafile. Eric Farr has a copy of the test program and has verifed that the problem exists. ═══ 2.210. 60857 beta: using find folder on desktop causes instablitiy ═══ name 60857 reference abstract beta: using find folder on desktop causes instablitiy duplicate defect 59634 state canceled severity 4 answer duplicate symptom fnwad duplicate defects: none. duplicate features: none. notes: Using RMB on desktop then select find and ALL selections down thru the file folders. Find window opens and then selections are displayed. Close the window and do it again. The second window will have nothing showing in it. Close the window and do it again. The 3rd and 4th etc,, will be like the first. Now try to shred the 2nd empty folder and mouse will become unstable until the desktop is no longer useable. Another observation is that the system slows - considerably - almost as if locked. this sounds like the same problem Both deal with find results folders hanging system. duplicate ═══ 2.211. 60935 BETA: OS/2 2.0 Desktop ═══ name 60935 reference abstract BETA: OS/2 2.0 Desktop duplicate state canceled severity 3 answer invunr symptom use duplicate defects: none. duplicate features: none. notes: The filename of the Desktop is "2.0" and not "2.1". A Beta Tester If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX. Shelia, this a dup of 59514. 59514 is closed and it went into yesterday's build. Duplicate of 59514 ═══ 2.212. 60936 BETA: DIR /R -- Incorrect Output ═══ name 60936 reference abstract BETA: DIR /R -- Incorrect Output duplicate state verify severity 3 answer chkm symptom use duplicate defects: none. duplicate features: none. notes: Subject: DIR /R -- Syntax Error If you issue the command "DIR /R" in an OS/2 session, you get "SYS1003" (a syntax error). If you issue the "DIR /?" command in an OS/2 session, the limited help implies that "/R" is OK. The command works fine in a DOS session. In the online Command Reference, the "/R" parameter is described (though not too clearly) but no mention is made of "works in DOS, not in OS/2 session." Either the online Command Reference should be updated, along with removing the "/R" description when doing "/?" in OS/2, or the DIR command should be fixed to work in OS/2 sessions. A Beta Tester If you have any questions, please contact Sheila McDonald-Rich on 3-XXXX. This is a DOS mode only feature and the documentation is being corrected to say that. Mark Rogalski ID will modify the online help to reflect the correct DIR syntax. Dir /R needs to be updated in the system messages to indicate that this is a DOS mode only parameter as it has been updated in the Online Command Reference per defect 59749 which this defect 60936 was originally dupped against. I have spoken with Anita regarding this change and she agrees that it is in component ID_SYSERROR. If any questions, please contact Sheila McDonald-Rich on 3-XXXX. Thanks Screen Team Approved ═══ 2.213. *60937 Word Perfect: objects with .classinfo won't change class ═══ name 60937 reference abstract Word Perfect: objects with .classinfo won't change class duplicate state closed severity 2 answer chkm symptom fnwad duplicate defects: none. duplicate features: none. notes: The attached was submitted by Chris Andrew (ex IBM) for WordPerfect. ----- NDOPC1 FORUMISV appended at 11:24:58 on 93/01/18 GMT (by IBMOS2 at BCRVMPC SUBJECT: NDOPC1 / N3571/MCI3169 A serious Workplace Shell problem 28162 --IBMMAIL Date and time 93/01/15 15 :02:36 Just discovered a show-stopper problem for WordPerfect that involes the Workplace Shell. I am writing code that will 'morph' an object from one object class to another object class by merely changing its .TYPE extended attribute. This function should really be provided by OS/2 - however I appreciate that it is a non-trivial problem to solve and have been creating a work-around that will run on all levels of OS/2 irrespective of whether this function is added to OS/2 2.0 or not.... HERE IS THE BUG: ~~~~~~~~~~~~~~~~~~~~ Workplace Shell allows an object class to register an association to a given .TYPE extended attribute. In theory, any file object that posesses that .TYPE extended attribute will become an instance of that associated object class. This works fine EXCEPT if the file has a .CLASSINFO extended attribute, when Workplace Shell thinks that it already knows the object class and erroneously decides to ignore any object class associations. In simple terms, this means that an object class association will take effect for all file objects out on your hardfile EXCEPT the ones that you touched using WorkplaceShell (eg: you added a new item to the menu or marked it as a template). Cheers, ChrisAn Extra X400 information begins: Originator Name: Christopher Andrew Organisation: CORP Domain: US/TELEMAIL/WPCORP Domain Attrib: WPMAIL:OS2(CHRISAN) Node.Userid: IBMX400.28162 Message Id: 2B57345E.B1F01 Importance: Normal Autoforwarded: No Sent by Name: Christopher Andrew Organisation: CORP Domain: US/TELEMAIL/WPCORP Domain Attrib: WPMAIL:OS2(CHRISAN) Node.Userid: IBMX400.28162 Subject: A serious Workplace Shell problem Recipients Name: BTIG BTI Domain: US/IBMX400/IBMMAIL Node.Userid: BCRVMPC1.OS2NDOP Report request: Yes Reply request: No please put descriptions of the problems in the abstract field. I can't close this out. What do I have to do? ═══ 2.214. *60951 BETA: Vdisk will not create 1024 sectors ═══ name 60951 reference abstract BETA: Vdisk will not create 1024 sectors duplicate state closed severity 4 answer doci symptom fnwad duplicate defects: none. duplicate features: none. notes: Documentation states that VDISK will create 1024 sectors. However the maximum seems to be only 512. Selecting HELP on the details page reflects little info on the notebook detail page and omits the word 'virtual' when using the notebook within the drive icon - notebook detail page using HELP. Reroute to right component Virtual disk. Miguel Asensio 2-XXXX. The code explicitly checks for the 1024 parameter and sets it back to 512. There is no documentation as to why this was done but it's believed to have been done as a performance enhancement. All system buffers will be as large as the maximum sector size used and, at 1024, this is a considerable waste of space. Instead of rejecting the parameter and failing to initialize, VDISK resets the parameter to 512 to allow user to boot and have a working VDISK. The documentation should be updated to document this change. Miguel Asensio 2-XXXX. cancelled - code can not be changed / documentaion should be updated to reflect limitation of 512. defect should be re routed to id_commandref so that documentation can be changed to reflect the restriction of 1024. Screen Team Disapproved Please see text for explanation of problem. Screen Team Approved ═══ 2.215. *60956 BETA: Set Text Alignment (GSTA) Documentation Error ═══ name 60956 reference abstract BETA: Set Text Alignment (GSTA) Documentation Error duplicate state closed severity 3 answer doci symptom vis duplicate defects: none. duplicate features: none. notes: A developer has reported the following problem. Prog. Ref. Vol III, page 33-47. Set Text Allignment shows that the metafile has 3 bytes to describe the allignment of text that follows. 0x36 0xHORIZ 0xVERT where 0xHORIZ and oxVERT are the horizontal and vertical alignment flags TA_*. However the hex dump of the metafile reveals 4 bytes describing text allignment. 0x36 0x02 0xHORIZ 0xVERT where 0x02 is the length of the data following the graphic order. Debbie Best has confirmed that the metafile is correct and that the documentation is in error and needs to be changed. Screen Team Approved ═══ 2.216. *60958 BETA: METAFILE Text Allignment Is Incorrect ═══ name 60958 reference abstract BETA: METAFILE Text Allignment Is Incorrect duplicate state closed severity 3 answer dsai symptom appf duplicate defects: none. duplicate features: none. notes: A developer supplied a test program TESTTXT.EXE that reproduces a problem with text allignment. Debbie Best has the test case. 1) Top to bottom allignment is incorrect. 2) When playing a metafile horizontal and vertical allignments are being reversed. When the metafile was played the GPI was reversing the horiz and vert parameters before calling the engine. This was fixed as was a problem in the GRE where the top and bottom alignments were reversed. ═══ 2.217. 60981 BETA : Quicken 2.0 for Windows does not install ═══ name 60981 reference abstract BETA : Quicken 2.0 for Windows does not install duplicate state closed severity 3 answer dsam symptom appf duplicate defects: duplicate features: none. notes: Quicken 2.0 does not install. Files olesrv.dll, olecli.dll, and ddeml.dll cannot be renamed. Lisa 2-XXXX Reassign Screen Team Approved The problem is with the Quicken install program. It uses OpenFile (windows API) without realizing that the PATH will be searched. Thus, when install tries to open A:\COMMDLG.DLL to see if it is on it's diskette, it actually opens the system copy of COMMDLG.DLL, etc. instead. Thus, when it tries to delete this file later, it can not because there is an open reference to it. The work-around is to rename the 3 system files before attempting to install Quicken. Mark Rogalski This app installs with no problem under m/s win 3.1 and is a very popular application. If it installs under win 3.1, it seems to me that it should install under win-os/2 3.1. Transferring per our discussion. Rerouted per George. The answer code is bogus... there is no valid choice that is accurate. ═══ 2.218. 60982 This may not be the file system, but swapper problems are usual ═══ name 60982 reference abstract This may not be the file system, but swapper problems are usual duplicate state verify severity 3 answer regr symptom fnwad duplicate defects: none. duplicate features: none. notes: This may not be the file system, but swapper problems are usually written against the file system, so........... if this is not the right place, please reroute this to the proper component. Runnning the beta. No problem. Got to a point where the swap file could not grow any further and I got the popup. I told the popup to kill the program causing the swap file to grow, just as I should. When the program goes away, and the popup goes away, my mouse is stuck in the top left corner of the screen. Easy to reproduce. Problem only goes away when you reboot Thanks. There is not enough information to determine this belongs to the swapper, let alone the file system. Even though this can be recreated it needs to be looked at further before a component is assigned. Please find the proper component for this before reopening. G. Simco Red, I have talked to Val about this defect. There is not enough information here to investigate the problem. Why did you reassign this to smg? Please give me a call when you can. Claudette This may not be the file system, but swapper problems are usually written against the file system, so........... if this is not the right place, please reroute this to the proper component. Runnning the beta. No problem. Got to a point where the swap file could not grow any further and I got the popup. I told the popup to kill the program causing the swap file to grow, just as I should. When the program goes away, and the popup goes away, my mouse is stuck in the top left corner of the screen. Easy to reproduce. Problem only goes away when you reboot Thanks. *********************************************************************** I have neither the time nor expertise to narrow this any further. We can throw it back and forth and leave it "unfixed" or we can work together to find the correct component. We will be shutting down soon and if we really care about the quality of the product we should not block each other's efforts. If this is not the correct component, I need help in identifying the correct one. Thanks. This may not be the file system, but swapper problems are usually written against the file system, so........... if this is not the right place, please reroute this to the proper component. Runnning the beta. No problem. Got to a point where the swap file could not grow any further and I got the popup. I told the popup to kill the program causing the swap file to grow, just as I should. When the program goes away, and the popup goes away, my mouse is stuck in the top left corner of the screen. Easy to reproduce. Problem only goes away when you reboot Thanks. *********************************************************************** The business with the session manager was an error on my fault. Sorry. I have neither the time nor expertise to narrow this any further. We can throw it back and forth and leave it "unfixed" or we can work together to find the correct component. We will be shutting down soon and if we really care about the quality of the product we should not block each other's efforts. If this is not the correct component, I need help in identifying the correct one. Thanks. BETA:User on OS2BETA (Deitmar@sydvm1,23:48:09 on 93/01/24) reports the same problem when he responds to ignore the SWAP PARTITION FULL message. He also states he could switch to a fu OS/2 PM desktop. At this point, the mouse pointer was freed again. This may not be the file system, but swapper problems are usually written against the file system, so........... if this is not the right place, please reroute this to the proper component. Runnning the beta. No problem. Got to a point where the swap file could not grow any further and I got the popup. I told the popup to kill the program causing the swap file to grow, just as I should. When the program goes away, and the popup goes away, my mouse is stuck in the top left corner of the screen. Easy to reproduce. Problem only goes away when you reboot Thanks. *********************************************************************** The business with the session manager was an error on my fault. Sorry. I have neither the time nor expertise to narrow this any further. We can throw it back and forth and leave it "unfixed" or we can work together to find the correct component. We will be shutting down soon and if we really care about the quality of the product we should not block each other's efforts. If this is not the correct component, I need help in identifying the correct one. Thanks. Glen Brew said to reassign this to MOUSE. Dave Reich has a system all set up to demonstrate this. Dave is in Bldg. 12, room W2326 and his extension is 2-XXXX I hate to tuck a note away at the bottom of this where no one will see, but the problem here is that the mouse driver believes it is still in text mode when it is on PM after a harderror. That's why it's trapped in the upper left hand corner. I'm looking into who missed telling the mouse driver that it's back on PM. Bernie Thompson 2-XXXX Somebody is not updating the mouse driver's data structure once the screen is switched back to PM from the harderr. Therefore, the mouse still thinks it is in text mode. I believe the problem is is bvs (OS2Char) routine NotifyMouse. Screen Team Approved This problem does not happen with VGA32 or SVGA .. it seems to be only happening on XGA. Rerouting to XGA32 .. The fgndsessn is being properly updated. 0 for the harderror, and back to 1 upon entering PM. The mode data structure still reports text mode, however, while PM has already come up. moving to XGA32 .. (sorry joe! :) The Mouse problem specified in the defect is caused by excessive usage of file handles under the PMShell process. When the PM session is in the mist of becoming foreground, the PM Display driver will call VioSetMode. The failure occurs when BVS calls MouOpen with a return code of 4, indicating an out of file handle condition. Looking at the BMS code, I noticed that DosOpen to Mouse$ completes fine but DosOpen to SingleQ$ failed. One solution is to close the Mouse handle before doing the DosOpen to SingleQ$ but I would first like to find out who is taking all the available handles. ═══ 2.219. *60998 self extracting dos app fails with dos=high,umb ═══ name 60998 reference abstract self extracting dos app fails with dos=high,umb duplicate state closed severity 3 answer chkm symptom use duplicate defects: none. duplicate features: none. notes: A self extracting application that a customer has sent to me created with ZIP2EXE from PKWARE will cause sys3176 when executed in a VDM with DOS=HIGH,UMB. Without this application works fine. Please contact me for a test case. Brian Duquette 2-XXXX Screen Team Approved This is actually new code and new function. ═══ 2.220. *61022 BETA: File IBM2M57.ADD placed in \OS2 on NON-IBM machines ═══ name 61022 reference abstract BETA: File IBM2M57.ADD placed in \OS2 on NON-IBM machines duplicate state closed severity 2 answer dsai symptom fnwad duplicate defects: none. duplicate features: none. notes: Created by CMVC to address defect 60063. The file IBM2M57.ADD for the multimedia machines is being placed in the \OS2 subdirectory by the installation even on NON-IBM machine types. assign to ins from inst Deleted IBM1M57.ADD at the end of install. Reopen due to reproduced error in 6.488 on 486-33 Micronics Mthrbrd. CMVC:Fixed incorrect AFLD2 field. Added ibm2m57.add to the list of files to be deleted in cleanup.c. IBM2M57.ADD is not not needed on any machine once it has been installed. ═══ 2.221. 61033 BETA : keyboard focus not available ═══ name 61033 reference abstract BETA : keyboard focus not available duplicate defect 59591 state canceled severity 3 answer duplicate symptom use duplicate defects: none. duplicate features: none. notes: Forgive me if this has already been reported as I think it may have been but cannot seem to find a defect for it and it continues to propogate to the latest builds. When using an OS/2 session( fullscreen or windowed, if one uses /p or | more to page a listing say a dir, and uses CTRL-C to break out before completion, the next time | more is used, one receives the message sys0447 the keyboard focus is not available and sys 1805 the process tried to write to a non- existant pipe. ═══ 2.222. *61048 BETA:OEMHLP$ BIOSINFO Error for Model-57 ═══ name 61048 reference abstract BETA:OEMHLP$ BIOSINFO Error for Model-57 duplicate state closed severity 3 answer dsai symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA:User submitted this problem on the OS2BETA forum. I checked with Paul before writing this defect. ---------------------------------------------------------------------- ----- OS2BETA FORUM appended at 12:12:04 on 93/01/25 GMT (by KENKAHN at PKSMRVM) Subject: $OEMHLP BIOSINFO Error for Model-57 When issuing a BIOSINFO (function = 5) call to $OEMHLP for a PS/2 Model-57, it incorrectly returns that ABIOS is *not* available. Since this is a MicroChannel machine, it should have ABIOS support. This is causing at least two programs I have to fail. Kenneth Kahn (8)293-XXXX Staff of IBM Fellow:IBM Personal/370 Code written to support RAM loaded ABIOS was not functioning correctly. Fixed. Paul. ═══ 2.223. *61066 Beta-IPMD receives a 303 return code from DOSDEBUG ═══ name 61066 reference abstract Beta-IPMD receives a 303 return code from DOSDEBUG duplicate state closed severity 3 answer intm symptom use duplicate defects: none. duplicate features: none. notes: Beta - IPMD receives a 303 return code from DOSDEBUG PROBLEM: To recreate the problem do the Following: Under sty_dlg, bring up the source for InitializeNotebook. Set a breakpoint on the closing bracket of the InitializeNotebook function. It should be line 1741 in sty_dlg. Run. Bring the application window up. Select Demo->Dialog controls->Cua Controls->Notebook. This will cause the breakpoint to be hit. Step over several times until you get to a call in PMWIN:1. The call instruction should be at D0DF:2943. Step into the call and continue stepping until you are on the 2nd MOV instruction past the CLI instruction. When you step over this MOV, a DosDebug Error popup appears, then a SYS1934 error. Solution: When DosDebug Severs the Connection do not return a return code. I have fixed the problem, we just need the Defect for the build. If you have questions or concerns please give me a call. Happy Debugging, Arthur Roberts Jr. T/L 443 - 3148 ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** *** *"A problem well stated is a problem half solved." * *************** By Charles Kettering ************************ ═══ 2.224. *61098 BETA: message sys0154I incorrect ═══ name 61098 reference abstract BETA: message sys0154I incorrect duplicate state closed severity 3 answer docm symptom use duplicate defects: none. duplicate features: none. notes: Message is incorrect and spoke with Anita Soto and she will Correct message SYS0154I. Anita 2-XXXX. Correct message SYS0154I. Anita 2-XXXX. ═══ 2.225. 61110 BETA: program traps on pmgre call ═══ name 61110 reference abstract BETA: program traps on pmgre call duplicate state returned severity 3 answer invinf symptom thipe duplicate defects: none. duplicate features: none. notes: SAS - Andy Barnhart - program is trapping on PMGRE calls. This is a major problem which they would like to get solved before the release of their new software to their customers in the next few days. It occurs in both the 2.0 release with service pak and in 2.1 Beta. Information from SAS's debugger: XCPT_ACCESS_VIOLATION (C0000005) read access at address 0000008a at 0x1ac52bb2 pmgre:\_UNKNOWN_MODULE\UNKNOWN at 0x1ac52bb2 eax=0000005a ebx=16efc2ac ecx=00000001 edx=00000100 esi=0015005e edi=0cb0447c ebp=00154044 eip=1ac52bb2 ef1=00012206 esp=00154010 ds=0053 es=0053 fs=150b gs=0000 cs=005b ss=0053 n6 re nn nv up ei pl nz na pe nc The line that is taking the hit: 0x1ac52bb2: CMP DWORD PTR (EAX+30),FF Dan Lee Jim, as discussed with Dan Lee, we are going to return for more information. Although we now have enough here to understand why the defect was written, we need to get a remote debug session going to understand if this is in the graphics engine or another part of the system. Let me know if there is something we can do to help expedite this process, because I know how important this defect is to SAS. JimmyG Screen Team Approved Jim and Dan, thanks for the note. Please return this defect to me when we have the information. Basically, we are looking for the name of the routine where we are trapping. If we can get a map file that matches their level we can maybe determine it from that. Also it would be interesting to know what they are running and if they had a circumvention, what it was. Thanks, Debbi 3-XXXX ═══ 2.226. 61112 BETA:PMChart Printer Select fails if >1 print object to LPTduplicate ═══ name 61112 reference abstract BETA:PMChart Printer Select fails if >1 print object to LPTduplicate state canceled severity 3 answer symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: If you have more than 1 printer object with an output to LPT! PM Charts Printer Setup will show only one printer and "none" . The printer displayed is not always the the default printer. If you choose the printer listed, you d o not allways get the printer setup notebook for that printer. (Re-entering remarks) Problem was determined to be a Mirrors problem. Larry Moore (IBM) said that it could be a Mirrors problem if PMChart is a Mirrors app . Steve Sachs (MicroGraphics Developer) confirmed it to be a Mirrors app, and investigated it and confirmed that the problems was Mirrors. Bill Jeter confirmed this cancellation. ═══ 2.227. 61118 BETA: DosSetSession wont keep session out of WindowList ═══ name 61118 reference abstract BETA: DosSetSession wont keep session out of WindowList duplicate state closed severity 3 answer chkm symptom fnwad duplicate defects: none. duplicate features: none. notes: This API worked in GA. I have a test case that starts the HELLO sample with a DosStartSession, then calls DosSetSession with the flag SET_SESSION_NON_SELECTABLE. In GA it worked: the DosSetSession API prevents the HELLO sample from appearing in the Window List. In the 2.1 beta, the session text "HELLO SAMPLE" does appear in the window list, but of course it shouldn't. This was opend for Borland, since they use DosSetSession in their debugger to prevent users from selecting debuggee sessions that may have breakpoints set in them. (Which would cause a system deadlock). Test case available. Dan Lee Screen Team Approved This defect was working as designed and never work in GA. This was a timing problem that occurred when a dossetsession was called to set the switch list block with the apps data and when the window was created it registered the switch list with the default settings. The dossetsession settings were overwritten. To fix this problem another switch list check was added to check for the PID in framesetswitchandtitle. Rerouting to pmcontrols Val ═══ 2.228. *61130 Shutdown fails with seamless win3.1 enhanced session ═══ name 61130 reference abstract Shutdown fails with seamless win3.1 enhanced session duplicate state closed severity 3 answer chki symptom thipe duplicate defects: none. duplicate features: none. notes: If one tries to shutdown with a seamless Win3.1 enhanced mode application running in the forground. Shutdown will not complete. It will hang after asking whether it is OK to close the application. To duplicate: -Migrate clock.exe -change the settings to run windowed and in 3.1 enhanced mode. -open up clock - leave it in the forground -shutdown Some interesting notes: 1) Shutdown does not fail if seamless sessions enhanced or standard are running in the background. 2) If one attempts first to shutdown with a standard mode application running in the foreground before attempting with an enhanced mode seamless session, shutdown completes fine. 2) If one attempts to shutdown with a standard mode application in the forground after attempting with an enhanced mode session, the standard mode session will also hang shutdown. Glenn and I have debugged this to a problem in PMDDE. Donald Hobern 2-XXXX Return codes from WinPostQueueMsg API were being interpreted incorrectly, subsequently causing us to kill a thread abnormally. Screen Team Approved ═══ 2.229. 61137 BETA: WINOS2 program area on WPS dead during load. ═══ name 61137 reference abstract BETA: WINOS2 program area on WPS dead during load. duplicate state canceled severity 1 answer invwad symptom vis duplicate defects: none. duplicate features: none. notes: Run a WINOS2 seamless session like Corel 2.5's WFNBOSS, select CONVERT ALL fonts. While WFNBOSS is converting fonts, move any window on top of WFNBOSS screen, then move window and imprint remains covering WFNBOSS. If WFNBOSS is minimized to the viewer the pointer is still a hourglass and you cannot select any Icon within this area. Example available in 13-W240 lab w/DEBUG terminal access. We have looked at this and it looks like a seamless windows problem. We tried calling Darren Miclette, but didn't reach him. If this is not you guys, sorry. --Sheila 2-8252 Reassign Screen Team Approved I admit that this is a usability problem, however, the seamless code is working as designed. Basically, Windows is a co-operative multi-tasker. Tasks only get cpu time when the currently running task gives up the cpu. That's why OS/2 is a better OS. It uses pre-emptive multi-tasking. In this particular situation, the WFNBOSS application is not relinquishing control of the CPU until is completes its font conversions. Thus, the tasks which handle painting of the app's window do not get to paint until this is done. However, the Winos2 session is only one of many running under OS/2. The rest of them will still get CPU time and therefore you can bring up the task list and request that an app be minimized or maximized. Thus this behavior should be expected from Windows apps which are not very co-operative. If you were to run this app in regular Windows, I believe you would always have the busy mouse prompt, and would not be able to take any action until the conversions are finished. Mark Rogalski Canceled Per Usability problem. This defect should go in the build because if the area on the desktop which was under a now minimized seamless Win app and you needed to open the OS/2 System icon to format a disk or move files with Drives icon you cannot get access to the icon therefore the System files. This can become more cumbersome for novice users. ═══ 2.230. 61144 beta: CD-ROM lock eject unlock has no indicator ═══ name 61144 reference abstract beta: CD-ROM lock eject unlock has no indicator duplicate state canceled severity 3 answer invwad symptom vis duplicate defects: none. duplicate features: none. notes: With a CD-ROM in the drive the LOCK EJECT UNLOCK panel is displayed - but there is no check mark next to LOCK or UNLOCK to show the status This is working as designed. Here is the explanation: When the context menu for the CD-ROM drive is brought up, we first try to query the status of the drive at that time, to see if it is currently locked, or currently empty, etc. Some device drivers do not support/provide this status information. Therefore, in the absense of this contextual information, we add all three (lock, eject, unlock) actions to the context menu. If you had a device that supported this status information then you would, for example, see only the unlock option if the drive was currently locked (because you can't eject or lock a locked drive). In any case, there wouldn't be a checkmark next to any of the three options, you would only see certain ones based on the current status of the drive (if that info is available). -Sheila 2-XXXX ═══ 2.231. 61145 BETA: CD-ROM drive background settings corrupts ═══ name 61145 reference abstract BETA: CD-ROM drive background settings corrupts duplicate state canceled severity 3 answer intmt symptom vis duplicate defects: none. duplicate features: none. notes: When setting the background color to other than the default - it will not stay - upon rebooting the default background color is used. NOTE>> On the 488 drop trying to set the background color allowed me to go into the settings notebook ONCE - then upon exiting I could not go into it again and I could not access the drive. The first part of this defect, that settings are not saved for CD drives across re-boots, is due to the fact that CD drives are read-only. Ideally, we should put up some sort of message when the settings notebook is opened for a read-only drive saying that the settings will not be saved after re-boot. We don't want to entirely disable settings because some people leave their machine powered on for long enough periods of time for the settings to be useful to them. We would like to consider this for the next release. As for the second part of your note, the error is caused by a divide by zero exception (already reported in defect 60064). and only shows up in the post 6.479c beta drivers. Please accept this to be trs-ed. --Sheila 2-8252 ═══ 2.232. 61228 CAT & Mouse About Box In Error ═══ name 61228 reference abstract CAT & Mouse About Box In Error duplicate state canceled severity 4 answer asnit symptom vis duplicate defects: none. duplicate features: none. notes: The "about box" for Cat & Mouse says "PMSeek" instead of Cat & Mouse. The copyright screen for cat and mouse says pmseek instead of cat and mouse. This screen is owned by development (the person who owns the code). Please route to development. Anita 2-XXXX. The copyright screen for cat and mouse says pmseek instead of cat and mouse. This screen is owned by development (the person who owns the code). Please route to Joe Tano's (development) area. Anita 2-XXXX. This fix requires a change in NEKO.DLG file which is a MRI file. If the translation can be done for the file, we can make the change. I understand that this will automatically get opened against the next release. ═══ 2.233. 61307 This is to reopen defect 60659 which was closed by me in error. ═══ name 61307 reference abstract This is to reopen defect 60659 which was closed by me in error. duplicate state canceled severity 3 answer invinf symptom thipe duplicate defects: none. duplicate features: none. notes: This is to reopen defect 60659 which was closed by me in error. Sorry. Why did you reopen this Red ? There was a code change for the last defect, that went into 6.490. Is there a problem with that change ? If so, what ? Returning user error - insufficient info. Paul. This is to reopen defect 60659 which was closed by me in error. Sorry. Paul, It wasn't/isn't clear to me from the text of 60659 that a fix had been made. It was in the verify state. This is all probably due to my lack of CMVC expertise. I just don't want the problem to get lost. I closed 60659, I'm not sure what I should do with it when it's in the verify state. My notes say my options are accept, reject or abstain all of which close the defect. If it is fixed, I wll close this one. Plz advise. Sorry for the confusion, just don't want to lose it. Returning for Red to sort out what he his doing. ═══ 2.234. 61308 BETA:IPL msgs. on DISKETTES - OS/2 !! SYS1475, OS/2 !! SYS2duplicate ═══ name 61308 reference abstract BETA:IPL msgs. on DISKETTES - OS/2 !! SYS1475, OS/2 !! SYS2duplicate state returned severity 3 answer docmp symptom vis duplicate defects: none. duplicate features: none. notes: If a formatted diskette( under 2.0 or greater )is left in the drive and the machine is rebooted, then the following messages appear: OS/2 !! SYS01475 OS/2 !! SYS02027 instead of a more descriptive message such as "non system diskette..." Screen Team Approved The text for these messages were removed for a more valuable finishing the closing action. The text for these messages were removed to provide a more valuable function. They are being documented in the 2.1 message file help function and in the installation guide. ═══ 2.235. 61312 BETA: ECHO statements in .CMD files are agonizingly slow ═══ name 61312 reference abstract BETA: ECHO statements in .CMD files are agonizingly slow duplicate state canceled severity 3 answer invunr symptom vis duplicate defects: none. duplicate features: none. notes: when executing .cmd files echo statements take between .5 and 1 second to execute. This poses quite a problem to people relying on .cmd files to perform routine tasks. Current driver does not exhibit the slow ECHO performance described by this defect ═══ 2.236. 61313 BETA: DIR /S \OS2 prints a 'K' as the 8th digit in 'bytes uduplicate defect 59550 ═══ name 61313 reference abstract BETA: DIR /S \OS2 prints a 'K' as the 8th digit in 'bytes uduplicate defect 59550 state canceled severity 3 answer duplicate symptom vis duplicate defects: none. duplicate features: none. notes: Type at a command prompt: DIR /S \OS2...The final result could be something like this: 1015 file(s) 4695506K bytes used 3022848 bytes free The 'K' in the 'bytes used' field should be a digit. Unreproducible - fixed by 59550 ═══ 2.237. 61314 BETA: DIR /S \OS2 prints a 'K' as the 8th digit in 'bytes uduplicate ═══ name 61314 reference abstract BETA: DIR /S \OS2 prints a 'K' as the 8th digit in 'bytes uduplicate state canceled severity 3 answer symptom vis duplicate defects: none. duplicate features: none. notes: Type at a command prompt: DIR /S \OS2...The final result could be something like this: 1015 file(s) 4695506K bytes used 3022848 bytes free The 'K' in the 'bytes used' field should be a digit. This is a duplicate...See defect 61313 ═══ 2.238. 61325 BETA:Wordperfect problem on Thinkpad 700c ═══ name 61325 reference abstract BETA:Wordperfect problem on Thinkpad 700c duplicate state canceled severity 4 answer intit symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA:Problem submitted from OS2BETA forum. Defect 59481 was previously opened for a related problem. ----- OS2BETA FORUM appended at 09:50:33 on 93/01/18 GMT (by PT04633 at Y ..... OS2BETA FORUM modified at 11:05:19 on 93/01/18 GMT (by PT04633 at Y Subject:Word Perfect Program Does not work correctly Ref: Append 01/18/93 17:32 The PS/2 with SVGA chip(Thinkpat 700C) has this problem The PS/2 with XGA or VGA chip has no problem. The PS/2 with SVGA chip with VGA display driver has no problem. The problem seems to be SVGA dispay driver in OS/2 2.1 beta === Word Perfect =========== On the main screen type several lines of text and vary the Size the font and underlining or insert a normal file into the program. Now perform a print preview by Shift F7 then 6 and you should see a black screen with the text in it, this is not not shown correctly, so press F7 followed by Shift F1 to bring up the configuration, now press 3 then 5 to list the screen drivers select the 640 x 480 16 colour VGA and press enter 3 times. Now press Shift F7 then 6 to perform print preview and you will get the message: Invalid Graphics Driver. Also when running any dos based program the screen will compress at the top and bottom. Thanks. H. Yasu for H. Kinoshita(JL05218 at Ymtvm6) Potable systems TRS this to next release This should be TRS'd to next release it involves outside Vendor ═══ 2.239. 61339 MSWORD 2.0b GPF in OS2k386.exe at 0001:4bd0 ═══ name 61339 reference abstract MSWORD 2.0b GPF in OS2k386.exe at 0001:4bd0 duplicate state closed severity 3 answer chkm symptom thipe duplicate defects: none. duplicate features: none. notes: MS Word for Windows ver 2.0b GP faults in OS2K386.EXE at 0001:4bd0 when invoked in seamless or full screen. Please call me, Brian Duquette at 2-XXXX, for a copy of this app. Brian, Please don't write any defects against Win-OS/2 before contacting me or another developer in Win-OS/2 3.1. There are pieces of Windows that are not owned by my group, and if the fault lies in one of these pieces then we simply have to resend it. Also, could you please include more information on how to recreate the problem? I suppose the lack of information simply means it doesn't run at all?! Thanks, Joe Hlavaty WinWord 2.0b was making an invalid call to GetCodeHandle(). Fixed it so they don't trap Win-OS/2. Screen Team Approved ═══ 2.240. 61348 Quicken 2.0 for Windows will not install ═══ name 61348 reference abstract Quicken 2.0 for Windows will not install duplicate defect 60981 state canceled severity 3 answer duplicate symptom appf duplicate defects: none. duplicate features: none. notes: If one attempts to install Quicken 2.0 for windows it refuses to install. It displays the message that it cannot overwirite current DDEML.DLL. If one renames this file, Quicken 2.0 will install normally without any apparent detriment to Winos2 due to the fact that the DLL was replaced. All apps appear to work fine and winos2 seems to work fine. This is a duplicate. ═══ 2.241. *61361 BETA: Verify in DOS full and window initialized to ON ═══ name 61361 reference abstract BETA: Verify in DOS full and window initialized to ON duplicate state closed severity 3 answer asnm symptom vis duplicate defects: none. duplicate features: none. notes: Documentation states that VERIFY is initialized to OFF. This is true in OS/2 window and Full screen sessions. It is NOT true in DOS window and Full screen sessions. Screen Team Approved ═══ 2.242. *61362 BETA: opn line help for Pciture viewer in Prod. folder ═══ name 61362 reference abstract BETA: opn line help for Pciture viewer in Prod. folder duplicate state closed severity 3 answer docm symptom vis duplicate defects: none. duplicate features: none. notes: If you click on Picture Viewer in the productivity folder to highlight it, and then press F1 - you get 'Help for Program Object' rather than a brief description of what Picture Viewer does. Working Anita Soto 2-XXXX. Screen Team Approved ═══ 2.243. 61370 BETA: This is similar to ═══ name 61370 reference abstract BETA: This is similar to duplicate state canceled severity 3 answer invinf symptom thipe duplicate defects: none. duplicate features: none. notes: Created by CMVC to address defect 60659. This is a re-opening of a problem that was fixed in defect 25316. The`symptoms are identical (according to the IBM developer in Rome (DCAF). I have a copy of the app and will get whatever detail is needed. I suspect that the fix got "lost" in r206. BETA: This is a re-opening of a problem that was fixed in defect 25316. Symptoms are identical (according to the IBM developer in Rome (DCAF). I have a copy of the app and will get whatever detail is needed. I suspect that the fix got "lost" in r206. This was written against mmm but the fix was applied to ifsm in r202, transferring to ifsm. Routing per discussion with Paul D. G. Simco Well the fix is in a different place, os it would seem to be a different problem. Paul. CMVC:Fixed incorrect AFLD2 field. Returning for Red to sort out what he his doing. ═══ 2.244. 61375 BETA: MS Access install causes General Protection Fault ═══ name 61375 reference abstract BETA: MS Access install causes General Protection Fault duplicate state verify severity 3 answer inti symptom vis duplicate defects: duplicate features: none. notes: When attempting to install Microsoft Access the following error is reported. STFSETUP caused a General Protection Fault in module STFSETUP.EXE at 0003:160F. Screen Team Approved J. Luu 61375 P100 01/12/93 MS ACCESS traps during install (1st diskette) This is due to the program made a call to GetModuleHandle(USER.EXE) and we return handle=0. A subsequence of GetProcAddress with handle 0, produces a GP fault. If we change OS2USER items in SYSTEM.INI to user.exe and rename OS2user to user.exe, then MS ACCESS is able to install. Solution. FindExeInfo only looking for filename. It can't handle filename extension. Therefore this fix insert a new section of code to scan for '.'. If there is one, then AX=strlen(ss:di) up to the period and FindExeInfo is called again to search for the EXE list items. File change LDAUX.ASM, function IGetModuleHandle Joe Luu 26843 ═══ 2.245. 61381 BETA: Shadow Icon Text color can not be edited ═══ name 61381 reference abstract BETA: Shadow Icon Text color can not be edited duplicate state canceled severity 4 answer intit symptom vis duplicate defects: none. duplicate features: none. notes: Color of the Icon Text of a SHADOW can not be edited. This problem has been reported long ago and TRS-ed. Please accept a TRS on this one, since we cannot dup a defect in the "returned" state. Note for the customer... a workaround is to do a system wide change (hold the alt key down) while dropping the scheme on the desktop, and that will change the shadow text font color (changes it everywhere, but it does change it). --Sheila 2-8252 ═══ 2.246. 61417 BETA: Incorrect message format when copying files. ═══ name 61417 reference abstract BETA: Incorrect message format when copying files. duplicate state canceled severity 4 answer invwad symptom vis duplicate defects: none. duplicate features: none. notes: An user reported this problem through the Internet...Here part of the text he sent us: "...There are a few small problems I encountered with OS/2 2.1 Beta which I will report as soon as I know who to report them to. Furthermore, I have a question or suggestion. OS/2's DOS still uses the syntax "2 FILE(S) COPIED." After 10 years of DOS, is it possible for IBM to change it in OS/2? So, copying 1 file would report in "1 FILE COPIED" and copying 2 or more files will result "2 FILES COPIED." This, of course, should work for any number of files. It's a small 'correction' but would enhanve OS/2 even further. Thanks." Reassign to proper owner and component Per Paul Cheatum this defect is being returned. It is a suggestion. Anita Soto 2-XXXX. Cancelling per Paul Cheatham's request. Screen Team Disapproved ═══ 2.247. *61445 BETA: WinQuerySendMsg, WinReplyMsg buggies ═══ name 61445 reference abstract BETA: WinQuerySendMsg, WinReplyMsg buggies duplicate state closed severity 3 answer asnm symptom thipe duplicate defects: none. duplicate features: none. notes: Opened for Mike Cooper: both of these API's (WinQuerySendMsg and WinReplyMsg) have thunk problems. routing to r205/r206 comp The private apis for PM debugging had a couple of thunk bugs - probably since the beggining of time. These have been fixed and now help prevent Borland from hanging in certain situations. Screen Team Approved ═══ 2.248. *61456 BETA: incorrect help msg for left hand right hand help ═══ name 61456 reference abstract BETA: incorrect help msg for left hand right hand help duplicate state closed severity 3 answer doci symptom vis duplicate defects: none. duplicate features: none. notes: Help message for the mouse incorrect. Verbage in Help for Right hand left hand usage needs correcting. Working Anita Soto 2-XXXX. Screen Team Approved ═══ 2.249. *61457 BETA: incorrect help msg for restore ═══ name 61457 reference abstract BETA: incorrect help msg for restore duplicate state closed severity 3 answer doci symptom vis duplicate defects: none. duplicate features: none. notes: Help message missing the word 'to' in the message Screen Team Approved We have corrected this error. Screen Team Approved ═══ 2.250. *61464 BETA: WinSetOwner allows bogus ownership chains ═══ name 61464 reference abstract BETA: WinSetOwner allows bogus ownership chains duplicate state closed severity 3 answer chkm symptom fnwad duplicate defects: none. duplicate features: none. notes: WinSetOwner will allow you to set a window to own itself. As in: hwndDlg = WinFileDlg( ... ); WinSetOwner( hwndDlg, hwndDlg ); //bogus!!! It will also allow ownership chains that look like this: owns owns +---> hwnd1 --------> hwnd2 ---------> hwnd3 ----+ | | +-------------<--------------------<-------------+ owns These circular ownership chains can cause WinQueryWindow to deadlock, and causes problems with WM_QUERYFOCUSCHAIN. Dan Lee Screen Team Approved Checking added for SetOwner and checking in XSetParent enabled. ═══ 2.251. *61481 Close the Desktop ??!! ═══ name 61481 reference abstract Close the Desktop ??!! duplicate state closed severity 3 answer dsai symptom use duplicate defects: none. duplicate features: none. notes: Here's something that could be considered a design issue, but I consider it a defect. The decision is based upon the question: When, if ever, should a user be allowed to close the desktop to the "dead" space beyond it. Never! Scenario: 1) open the drives folder. 2) open the settings notebook for the drive OS/2 is installed on( where the desktop folder resides ). 3) Click on the menu tab. 4) Click on OPEN in the Available menu's box. 5) Click on settings button that becomes selectable 6) Select Tree as default view 7) Click on OK 8) Close the settings notebook 9) Shutdown On reboot the desktop will be a windowed tree view because all folder's in the drive inherit the tree view as default. Then if one( picture a novice user ) closes the window the desktop appears in, you've entered the "Dead zone". This is an issue of should this be possible. At the very least the desktop window should have no close option. At the very most the desktop should not default to a window. Screen Team Approved now that subdirectories inherit their parent directory's open view default, we need to override for the desktop so it doesn't open (on boot) in any other view than icon. ═══ 2.252. 61507 Beta: Lotus Notes - Lost text in dialog boxs ═══ name 61507 reference abstract Beta: Lotus Notes - Lost text in dialog boxs duplicate state canceled severity 3 answer invextsw symptom vis duplicate defects: none. duplicate features: none. notes: OS/2 Beta: 6C.479 System: Mod 95 Memory: 16 Meg Display: XGA 8515 Application: Lotus Notes 112.0 Problem Description: This application consists of 14 diskettes that will create install files for Lotus Notes for DOS, Windows, and OS/2. When creating the OS/2 install diskettes after the 8th diskette the text in the "Installing Notes" dialog box is no longer visible. To reproduce the problem do the following: Note: This requires about 13 formatted diskettes. 1. After installing the program into a subdirector, Type the following on the command line: INSTPM.EXE -R -R and press enter. 2. Select OK. 3. Select "Create OS/2 Workstation Diskette Kit". 4. Insert formatted diskette in driver A:. 5. Select OK. 6. Select OK on the "Installing Notes" dialog box. 7. Notice that the "%" sign in the "Transferring Files" dialog box is clipped and partially visible. This is totally visible on GA. Notice all the text in the "Installing Notes" dialog box is visible. 8. Continue following the instructions for creating the installation diskettes. 9. Note on the 3rd screen of "Transferring Files" the fonts change in the field called "This disk:". The "%" sign is now visible but the in the "Installation:" field, the "%" sign is clipped. 10. The "This disk:" field keeps changing fonts. 11. After the 8th diskette, the "Installing Notes" dialog box has no text other than the title bar visible and it never recovers. This works on the GA code. I have the app to reproduce the problem. Bob Gonzalez 982-XXXX Bob, JimmyG said you were going to verify this still occurs on the latest borg driver. Please return it to us if you find the problem still exists. Thanks, Debbi. I ran the app on Borg 6.490 and the text in the dialog box miss has been corrected. The only problem left is that the '%' sign in the disk field is clipped the fonts in this field seems to change sizes. Screen Team Approved Bob, as you requested I've sent you a vm note explaining my findings on this problem. I have tested this defect and missing text problem has been fixed. Gonzalez ═══ 2.253. 61509 ═══ name 61509 reference abstract BETA: PUBPB for OS/2 gets memory error loading BMP, lock meduplicate state canceled severity 3 answer symptom appf duplicate defects: none. duplicate features: none. notes: Publishers Paintbrush for OS/2 from Zsoft, has a memory error loading large BMPs or GIFs. Memory sometimes locks and OS/2 halts. The second file loaded of any type causes error in memory. I have a copy of the software and a debug terminal! Sorry if this doesn't belong to kernel..I did not know were else to send it. Screen Team Approved Canceled Per Virtual Memory Function of PUBPB in error, defect will become wpshl comp. ═══ 2.254. 61516 BETA: color palette needs to change color 'system wide' ═══ name 61516 reference abstract BETA: color palette needs to change color 'system wide' duplicate state canceled severity 4 answer intit symptom vis duplicate defects: none. duplicate features: none. notes: Color Palette system wide change does not operate the same as FONT palette. In Color Palette taking the paint brush to the desktop -only changes the desktop. If brought to the windows then all windows change but not the desktop. FONT changes the desktop and all the windows. Consistency is what is needed. The problem is that when you drop a color on the desktop, it changes the background color setting for the desktop, and there is only one desktop. When you drop a color on a window, it changes the window color setting, which is different than the background setting for the desktop. There is only one font setting. So a system wide font change changes the fonts on the desktop and all windows. This has always worked this way, and probably needs to be changed so it is more consistent. This was not approved by the PTR reviewers as a ship issue for Borg. We will TRS it for now and consider for the next release. -Sheila 2-XXXX done in a later release ═══ 2.255. 61518 BETA: SETBOOT /Q ═══ name 61518 reference abstract BETA: SETBOOT /Q duplicate state verify severity 3 answer docm symptom use duplicate defects: none. duplicate features: none. notes: Subject: SETBOOT /Q In the on-line Command Reference, if you look at the description of the "/Q" parameter for SETBOOT, it tells you: Queries the currently set startup environment. Returns the default partition alias, timeout value, mode and drive letter assignments for unattended operation. To query the current startup environment, type the following: SETBOOT /Q Well, here's what *really* happens (cut from an OS/2 Window): ЫC:\|setboot /q Timeout 10 System Index 0 Mode advanced System 0 None System 1 None System 2 None System 3 None System 4 None I don't know if the documentation or the command is wrong, but the "mode and drive letter assignments" would be nice. A Beta Tester For additional info, please contact Sheila McDonald-Rich on 3-XXXX. Do Not use the IDEV component. It is the parent component for all the toolkit information. It has nothing to do with any ship-group documention. I spoke with the Command Reference writer and he says the information matches the code. If there is a prob lem with that, write the defect against the code. This is being reassigned because it does not appear to a documentation problem. This defect should go against the Command Reference. I believe this parameter is working as designed. The mode is normal/advanced the drive letters would be listed opposite the system IDs if there were any to list. Vince Spadea ID 2-XXXX Vince, Per my conversation with Jamie, the documentation for Setboot /Q is incorrect and needs futher clarification. He will call you with further information. Thanks, Sheila McDOnald-Rich Screen Team Approved ═══ 2.256. *61543 BETA:Names with period on HPFS system create duplicate foldduplicate ═══ name 61543 reference abstract BETA:Names with period on HPFS system create duplicate foldduplicate state closed severity 3 answer dsai symptom vis duplicate defects: none. duplicate features: none. notes: BETA: If you create a folder with a name that includes a period in the last position on an HPFS system, you wind up with duplicate folders. This problem was found by Mike Turner (2-XXXX) and seen by Mindy Pollack Screen Team Approved ═══ 2.257. 61549 BETA: Presparams not carried to PS ═══ name 61549 reference abstract BETA: Presparams not carried to PS duplicate state canceled severity 3 answer invwad symptom fnwad duplicate defects: none. duplicate features: none. notes: The best way to describe this problem is with an example. Suppose you have a dialog template that looks like this: DLGTEMPLATE ID_DIALOG BEGIN DIALOG "sample dlg", ID_CONTROL, 10, 10, 100, 10 PRESPARAMS PP_FONTNAMESIZE "8.helv" PRESPARAMS PP_BACKGROUNDCOLORINDEX CLR_DARKBLUE PRESPARAMS PP_FOREGROUNDCOLORINDEX CLR_WHITE BEGIN ...... END END In your dialog the text will be drawn in the correct font and colors. The correct font is used because the system applies the font to the PS before returning from WinGetPS or WinBeginPaint. The colors are correct, however, because the window receives a WM_PRESPARAMCHANGED for colors, and the color is remembered in the instance data. You can see the difference if you use a piece of code in your dialog routine that looks like case WM_INITDLG: hps = WinGetPS( hwnd ); clrBackground = GpiQueryBackColor( hps ); //color GpiQueryFontMetrics( hps, sizeof( FONTMETRICS ), &fm ); //font WinReleasePS( hps ); then check the value of "clrBackground". It will always be CLR_DEFAULT, not the color you specified in the dialog template. The font metrics call, however, returns the correct font that was specified in the dialog template. *All* presentation parameters for a window should be applied to a PS before the PS is returned to the caller. Dan Lee CLR_DEFAULT does not itself represent any particular color. It will mean that the system should use the current default for the window, whatever it is. That means that an application can set the color to CLR_DEFAULT and get the original default background color for THIS window without having had to keep track of presentation parameters, etc. There may be an asymmetry in the way that this works when compared with the font code, but the return of CLR_DEFAULT as the current color is correct. Donald Hobern 2-XXXX Hmm, touche' and all that. I added a WinSetPresParam to my code and used the default value I got to change the color of the text to the "default" background color. My text dutifully disappeared, meaning that the CLR_DEFAULT now equals CLR_DARKBLUE as set in the dialog box routine. Sorry for the trouble... (sheepish grin). Dan Lee 2-XXXX ═══ 2.258. *61553 BETA:Picview-Reading Metafile Performance Has Degraded ═══ name 61553 reference abstract BETA:Picview-Reading Metafile Performance Has Degraded duplicate state closed severity 3 answer dsam symptom prf duplicate defects: none. duplicate features: none. notes: Sent customer supplied file BIGMET.MET to Debbie Best. I am not sure if this is the component with the problem, if not please reroute to the right component. I will append a remarks file with additional info. From: V$IBITG7--BCRVMPC1 Date and time 02/02/93 17:23:54 To: SIERRA --BCRVMPC1 John M. Sierra cc: ANDRESS --BCRVMPC1 Andress, F.B. From: Mike Jasinski Subject: NO SUBJECT John, I talked with Alex and he said that Jeff Berkowitz had tested and verified that this problem in fact exists. Talked to Jeff and he said that he provided you with the information and that your group was investigating to see what could be done. Have you opened a defect yet? If not, I am willing to open the defect if you will tell me what component to write it against. I realize that you might not be able to fix it if it's related to the 32 bit graphics engine. But I would like to open the defect so that we can at least document it. Please advise. Thanks, Mike Jasinski Boca Technical Interface Group tl: 443-XXXX 407 443-XXXX 407-982-XXXX FAX *** Forwarding note from V$IBITG7--BCRVMPC1 01/29/93 12:43 *** To: ALEX --BCRVMPC1 Alex H. Tarpinian From: Mike Jasinski Subject: Alex, Please stay on top of this one. I will need a defect number when it's finally opened. Thanks, Mike Boca Technical Interface Group tl: 443-XXXX 407 443-XXXX 407-982-XXXX FAX *** Forwarding note from ALEX --BCRVMPC1 01/29/93 10:30 *** To: V$IBITG7--BCRVMPC1 From: Alex Tarpinian ***** IBM Internal Use Only ***** Subject: Metafile performance has degraded since GA. Depending on the machine type, the customer can see between a 1 to 8 second difference in speed. We are currently investigating the problem--no defect has yet been opened. Alex Alexander H. Tarpinian OS/2 System Certification, Release Lead/Printer Test Lead T/L 982-XXXX, ALEX @BCRVMPC1 *************************************************************** *** Forwarding note from V$IBITG7--BCRVMPC1 01/20/93 16:15 *** To: ALEX --BCRVMPC1 Alex H. Tarpinian From: Mike Jasinski Subject: ----- OS2BETA FORUM appended at 18:16:44 on 93/01/19 GMT (by REF at BCRVM7) --- Subject: Metafile performance Our application, among other things, displays large metafiles built from CAD drawings, and naturally people want it to be faster. When I compare 2.1 Beta (479) to 2.0 GA, using PicView to display the metafiles, I find 2.1 about 40% slower on identical hardware than 2.0. (e.g. 24 seconds vs. 16). (This is on mod 80 w VGA). Other potential variables should be pretty constant, as I'm booting the two OS versions from different partitions of the same box. Anyone else seen this? Bob French Boca Technical Interface Group tl: 443-XXXX 407 443-XXXX 407-982-XXXX FAX This defect is being returned as our performance department needs to provide some performance runs to show why (and where) this performance degradation exists. Mike, we expect to have this from them on Monday or Tuesday and will let you know when it arrives. JimmyG Per my conversation on 2/09/93 with Jim Giliberti, I am reopening this defect. Scott has the rts results and is investigating... Screen Team Approved The performance problems are due to the number of times memory is suballocated when handling regions. ═══ 2.259. 61569 Folder Icon Template has duplicate short cut key assignment ═══ name 61569 reference abstract Folder Icon Template has duplicate short cut key assignment duplicate defect 24584 state canceled severity 4 answer duplicate symptom fnwad duplicate defects: none. duplicate features: none. notes: I am not sure if this is the right component to open this against, if not please reroute this to the right component. Thanks. In the "Templates Folder" select the "Folder Icon", open settings then use the mouse to select the pop down menu from the left edge of the title bar. The popup menu for folders shows S as the shortcut key for both "Create a _s_hadow" and for "_S_ort". Needless to say, this doesn't work too well. I use Sort all the time, but rarely create a shadow. This is true for all context menu's, (i.e. just bring up the desktop's context menu). This has been reported a while back and is a dup of 24584. -Sheila 2-XXXX Screen Team Disapproved ═══ 2.260. 61581 BETA: Borland debugger (again), trashing display ═══ name 61581 reference abstract BETA: Borland debugger (again), trashing display duplicate state canceled severity 3 answer valunr symptom vis duplicate defects: none. duplicate features: none. notes: Debugging the "sort" sample app, they set a bp on line 562 in the "CalcThread" subroutine (which spawns off threads to do the sorting). If they stop at this bp in 2.0 the system hangs - they have a thread that is running a child process, and is waiting for a new command. It has processed a "stop" and has sent a message to the UI thread that it has stopped. The UI thread calls WinFillRect and hangs there. This seems to indicate that a display semaphore is being held, and as soon as anything is done there the system freezes. Now the good news.... on 2.1, the system does NOT freeze, but the display get hosed, things get drawn in the wrong places, in the wrong colors, etc, and a reboot is required. Dan Lee Screen Team Approved Eric from Borland reported this as no longer reproducible. ═══ 2.261. 61597 BETA: NO help for Set IPF_KEYS in command ref ═══ name 61597 reference abstract BETA: NO help for Set IPF_KEYS in command ref duplicate state canceled severity 4 answer invsug symptom vis duplicate defects: none. duplicate features: none. notes: No help for IPF_KEYS in the command reference online docs The SET command is used in many CONFIG.SYS statements to setup the Workplace shell environment. Documentation of all these statements is, given the Borg schedule, a subject for another release. Vince Spadea ID 2-XXXX Screen Team Disapproved done in a later release ═══ 2.262. 61601 BETA: BitBlt mess-ups in 256 color mode ═══ name 61601 reference abstract BETA: BitBlt mess-ups in 256 color mode duplicate state canceled severity 3 answer invwad symptom vis duplicate defects: none. duplicate features: none. notes: I couldn't find anything wrong with the code sample I was sent. What they're doing is stretching a bitmap that is being blted from a memory PS to a screen PS in 256 color mode. What it does to the bitmap is take something that looks like a nice photograph and change it so that the colors are all haywire, even thought the image outlines are still correct. The call scenario goes like this (in case you see something I missed...) ... GpiSelectPalette( hps, pImage->hPal ); WinRealizePalette( hwnd, hps, (PULONG)&ulRet ); pImage->bmp.cbFix = sizeof( BITMAPINFOHEADER2 ); GpiQueryBitmapInfoHeader( pImage->hbm, &pImage->bmp ); GpiSetBitmap( pImage->hpsMem, pImage->hbm ); aptl(0).x=0; aptl(0).y=0; aptl(1).x=rcl.xRight; aptl(1).y=rcl.yTop; aptl(2).x=0; aptl(2).y=0; aptl(3).x=pImage->bmp.cx; aptl(3).y=pImage->bmp.cy; GpiBitBlt( hps, pImage->hpsMem, 4L, aptl, ROP_SRCCOPY, BBO_IGNORE ); GpiSetBitmap( hps, 0L ); .... I have a demo program from COMMIX Software that displays the bitmaps, and exhibits the strange behaviour in response to the "stretch image to fit frame" menu selection. Dan Lee 2-XXXX Screen Team Approved This turned out to be a testcase bug. The customer has found his problem and corrected it. Joe Celi ═══ 2.263. *61665 BETA: wpclsmgr.sc should be shipped with 2.1 toolkit ═══ name 61665 reference abstract BETA: wpclsmgr.sc should be shipped with 2.1 toolkit duplicate state closed severity 3 answer lfm symptom use duplicate defects: none. duplicate features: none. notes: The Workplace Shell object class "WPClassmanager" does not have its .SC file provided with the toolkit. this means that it is (almost) impossible to successfully create a subclass of that object class, unless you fake the SOM compiler out with a fake .SC file that looks as though it defines a class called "WPClassManager". With WPCLSMGR.SC it will be possible to add functions to the SOMClassManager object on the Workplace Shell process. Dan Lee 2-XXXX Screen Team Approved Added wpclsmgr.sc and wpclsmgr.h to the toolkit layout. Rob Semonis 3-XXXX ═══ 2.264. *61669 BETA: Migration of ATM.INI does not preserve existing fontsduplicate ═══ name 61669 reference abstract BETA: Migration of ATM.INI does not preserve existing fontsduplicate state closed severity 3 answer dsam symptom vis duplicate defects: none. duplicate features: none. notes: During migration when asked to preserve exisiting fonts the install program disregards the request and writes over the existing font file. This problem is reported from OS2BETA CFORUM. Here is how I recreate the easy way. 1).Obtain the OS/2 2.1 Beta on CD-ROM. 2).Install OS/2 2.1 first time with format option, install with all features selection and copy standard Windows system. 3).Start Winos2, run ATM and install fonts. What we expect is \OS2\MDOS\WINOS2\ATM.INI, section [fonts] contains a bunch of font description and where it locates. Now save your ATM.INI so later we can compare it with after-reinstall. 4). Reinstall OS/2 Beta, this time without format, but all feature install. Now select Preserve existing Windows system option. 5). After install process is completed, rebooted, and check that ATM.INI is not preserved. Joe Luu 26843 Screen Team Approved ═══ 2.265. *61677 BETA: in HPFS file system extensions incorrect (COM,EXE) ═══ name 61677 reference abstract BETA: in HPFS file system extensions incorrect (COM,EXE) duplicate state closed severity 3 answer chki symptom vis duplicate defects: none. duplicate features: none. notes: File extensiong name of .COMxxxx and .EXExxxx are seen the same as .COM and .EXE extensions. COMxxxx and EXExxxx should be seen as DATA files while .COM and .EXE seen as program files and the associated icon should represent them as data or program icons. Screen Team Approved ShlQueryIconGivenFEAL2 was comparing only the first 3 letters of the extension to see if it was .CMD, .COM, .EXE, etc. With HPFS files, such as FILE.EXEDATA, we now treat this as a data file instead of an executable. ═══ 2.266. *61716 DOS_AUTOEXEC can hang VDM creation ═══ name 61716 reference abstract DOS_AUTOEXEC can hang VDM creation duplicate state closed severity 3 answer inti symptom use duplicate defects: duplicate features: none. notes: Entering non-valid filespec characters into the DOS_AUTOEXEC setting can hang the creation of a VDM, without any explaination as to why you are hung. With parameter passing, it is very likely to occur in the field. This will also require a cooresponding change to the default CONFIG.SYS SHELL= statement. Screen Team Approved ═══ 2.267. 61729 BETA: Borland debugger problem with DosStartSession ═══ name 61729 reference abstract BETA: Borland debugger problem with DosStartSession duplicate state canceled severity 3 answer invinf symptom use duplicate defects: none. duplicate features: none. notes: DosDebug doesn't notify program if TermQ field is used in a DosStartSession call. I discussed this problem with Dan, He will send the problem back to development went he gets more information from borland. Until then this is a place holder for upcomming work. Please direct question to Dan or myself. When I requested detailed test cases from Borland, they responded that this issue was not as important as others they are working on, and that they would forward me test cases when they had time to pursue this problem. I'll re-open this when they let me know they're ready to go back to work on it. Dan Lee ═══ 2.268. 61730 BETA: Borland debug problem with Exit list in DLLAPI sampleduplicate ═══ name 61730 reference abstract BETA: Borland debug problem with Exit list in DLLAPI sampleduplicate state canceled severity 3 answer invinf symptom use duplicate defects: none. duplicate features: none. notes: DosDebug missed when DosExist called from Exception Handler in DLLAPI sample program. I discussed this problem with Dan, He will send the problem back to development went he gets more information from borland. Until then this is a place holder for upcomming work. Please direct question to Dan or myself. When I requested detailed test cases from Borland, they responded that this issue was not as important as others they are working on, and that they would forward me test cases when they had time to pursue this problem. I'll re-open this when they let me know they're ready to go back to work on it. Dan Lee ═══ 2.269. 61733 Create Another - Printer - new name ═══ name 61733 reference abstract Create Another - Printer - new name duplicate state open severity 3 answer symptom vis duplicate defects: none. duplicate features: none. notes: When you use the Create Another option from a printer objects pulldown menu to create another printer, the first window allows you to rename the object. However, when you select create the new name does not propogate to the next window. This does not happen with other objects because this second window is specific to the printer object Inside wpprt\wppdlg1.c, in procedure DisplayPrintDestNew, the text string for the printer name is loaded from the STR_PRINTMANAGER string in the res file. I don't believe we currently have an interface for passing the new name typed in on our dialog to the printer driver dialog. -Sheila 2-XXXX Screen Team Approved ═══ 2.270. 61750 BETA: Setting new DOS_AUTOEXEC looses command.com path ═══ name 61750 reference abstract BETA: Setting new DOS_AUTOEXEC looses command.com path duplicate defect 61716 state canceled severity 4 answer duplicate symptom vis duplicate defects: none. duplicate features: none. notes: In a DOS session - using DOS SETTINGS and changing DOS_AUTOEXEC to look at different AUTOEXEC.BAT file causes COMSPEC to change in the session from C:\OS2\MDOS\Command.Com to \Command.Com Some day people will learn what component to write defects against. This is already being addressed. ═══ 2.271. *61751 BETA: Extra characters inserted when using ATL-L mouse to eduplicate ═══ name 61751 reference abstract BETA: Extra characters inserted when using ATL-L mouse to eduplicate state closed severity 3 answer dsai symptom vis duplicate defects: none. duplicate features: none. notes: Usage of ALT-LEFT MOUSE keys to edit an object's title results in extra characters added to the end of the title. To reproduce the problem, hit the ALT-LEFT MOUSE on an object; change the title and hit enter. Then open the Settings notebook, go to the General tab and the object's title has music notes at the end of the string. The music notes can be edited and removed. However, if the object's title is edited through the Settings notebook, these music notes are not added. Screen Team Approved change MLM_FORMAT from MLFIE_NOTRANS to MLFIE_CFTEXT, so carraige returns won't show up as graphical chars. ═══ 2.272. 61752 BETA: Left-handed mouse configuration is not working properduplicate ═══ name 61752 reference abstract BETA: Left-handed mouse configuration is not working properduplicate state verify severity 3 answer chkm symptom vis duplicate defects: none. duplicate features: none. notes: 1) Configure the mouse for left-handed operation 2) Go to the Icon Editor. Notice that the mouse is not reflecting the change made. If you use the right mouse - in the Icon Editor - you will see that it is using the wrong color and vice versa. This defect requires a code change. Not for Information Development. Routing to appropriate component. Screen Team Approved The WPS is handling the switching of the mouse buttons correctly but the ICON Editor needs to swap the color values of the mouse icon that represents the selected colors by viewing them on the buttons of the mouse icon. Gordon Webster 3-XXXX The editor was failing to check the system value for inverted pointing device (left-handed mouse). It needs to switch the colors when drawing the image of the mouse in the status area whenever the left-handed mouse is in effect. ═══ 2.273. *61772 BETA: changing icon text color of details view isn't saved ═══ name 61772 reference abstract BETA: changing icon text color of details view isn't saved duplicate state closed severity 3 answer dsai symptom vis duplicate defects: none. duplicate features: none. notes: Open a details view of the desktop as well as the color palette. Change the color of the desktop details view by CTRL- RMB a color onto the details view. If you then close and reopen the details view the color has not been saved. When you reopen the details view, the column titles of the details view show up in the desired color, all other text comes up black. Screen Team Approved The fix for this requires a change in the container code. It never worked right in GA either. CnrRefreshPresParam doesn't properly handle the ulPresParm == 0 case. ═══ 2.274. 61777 BETA: Diskettes 1 thru 19 only? ═══ name 61777 reference abstract BETA: Diskettes 1 thru 19 only? duplicate state verify severity 3 answer chki symptom use duplicate defects: none. duplicate features: none. notes: Subject: Diskettes 1 thru 19 only? During installation, the user is told that diskettes 1 through 19 are required to complete the installation. I think mention of the Printer and Display Driver Diskettes should be made, too. A BETA Tester If you need further assistance, please contact Sheila McDonald-Rich on 3-XXXX. Screen Team Disapproved Removed lines that refer to the diskettes you will need. Screen Team Approved ═══ 2.275. 61780 BETA: Install All Features - Missing An Alternative ═══ name 61780 reference abstract BETA: Install All Features - Missing An Alternative duplicate state open severity 4 answer symptom use duplicate defects: none. duplicate features: none. notes: Subject: Install All Features - Missing An Alternative If you select the "Install all features" option, you cannot do the minimal alterations to the OS/2 and DOS configurations you can perform if you take the selective approach (things like changing the swap path, threads,etc.). Perhaps the pull-downs that allow modification should be moved "forward" one screen? A BETA Tester If you need further assistance, please contact Sheila McDonald-Rich on 3-XXXX. Screen Team Disapproved ═══ 2.276. *61781 BETA: DISkCOPY & DISKCOMP Examples ═══ name 61781 reference abstract BETA: DISkCOPY & DISKCOMP Examples duplicate state closed severity 3 answer docm symptom use duplicate defects: none. duplicate features: none. notes: Subject: DISKCOPY & DISKCOMP Examples I've seen the question asked dozens of times "how can I make a copy of my 3.5" diskette when I have only one?" (Substitute 5.25" for 3.5" for a common variation of that question.) Of course, all of us "pros" know that you can "DISKCOPY A: A:" or "DISKCOPY B: B:" to make copies of diskettes with only one drive (and can also use the DISKCOMP command to compare diskettes). What's missing from the on-line Command Reference, in both the commentary as well as the examples, is a simple statement that you can do this and the system will prompt you for "source" and "target" as appropriate. A BETA Tester If you need further assistance, please contact Sheila McDonald-Rich on 3-XXXX. Screen Team Approved ═══ 2.277. *61793 BETA:SYS0447 and SYS1805 on DIR /P, ctl-break,type xx.xx ! duplicate ═══ name 61793 reference abstract BETA:SYS0447 and SYS1805 on DIR /P, ctl-break,type xx.xx ! duplicate state closed severity 3 answer dsai symptom fnwad duplicate defects: none. duplicate features: none. notes: Created by CMVC to address defect 59591. BETA: Problem reported by user and reproduced by myself-Use os/2 window Do a DIR /P for a directory that is more than one screen full of data. When it pauses, do a ctrl-break. Right after the more shows up at bottom of screen, you get SYS0447-The keyboard focus is not available SYS1805-The process tried to write to a nonexistent pipe. Please transfer this to correct component if I have indicated the wrong one. There is an easier way to reproduce this problem. Simply hit CTRL-BREAK at the command prompt (that's right, at the [C:\] prompt), then try "MORE < CONFIG.SYS". The SYS0447 error will occur after one pageful of data, and the MORE command will not work at all for the remainder of the session. This problem occurs in both the OS/2 FullScreen and OS/2 Window. -- Kevin J. North, LAN Lab 3-2805, KNORTH@BCRVMPC1 Possible duplicate of 59536. The problem was that the per session semaphore was set as owned by KbdCharIn() and never cleared under the following conditions. - Run a full screen protect mode session - Hit - Use any program that makes keyboard API calls CMVC:Fixed incorrect AFLD2 field. Screen Team Approved OPPS... I didn't take into consideration the situation where a user could hit at the command prompt before piping to MORE.EXE. This caused the fast safe ram semaphore cUsage or count field to be out of synch with the KbdCharIn call from CMD.EXE. Gordon Webster t/l 443-XXXX ═══ 2.278. 61811 BETA: 256 color bitmaps display wrong as bkrnd to OS/2 deskduplicate ═══ name 61811 reference abstract BETA: 256 color bitmaps display wrong as bkrnd to OS/2 deskduplicate state canceled severity 3 answer intit symptom vis duplicate defects: none. duplicate features: none. notes: Have 256 color windows bitmaps from Columbia Univeristy. They display perfectly in Win/OS2 paintbrush running seamlessly and full screen. When used as the backround to the OS/2 desktop they do not display the proper colors. Screen Team Disapproved We would like to TRS this for now. We are placing it as a high priority in our "fit and finish" defects to possibly bring into borg later. -Sheila 2-XXXX ═══ 2.279. 61812 BETA: Startup Folder Contents overlay LOCK ON STARTUP BMP ═══ name 61812 reference abstract BETA: Startup Folder Contents overlay LOCK ON STARTUP BMP duplicate defect 59344 state canceled severity 3 answer duplicate symptom vis duplicate defects: none. duplicate features: none. notes: Select LOCK ON STARTUP with a BITMAP. Put a program template with progman.exe as the file in the Startup folder. Shutdown and then reboot. The Program Manager will overlay the LOCKUP bitmap but the user can do nothing until the password is entered. Screen Team Approved Mark Rogalski and I have looked at this problem and the problem stems from seamless having to paint over their black hole regardless if lockup is there or not. They do their set focuses which fail because we are sysmodal, but they paint anyhow. Since we really haven't lost focus and pmwin never told us we did, there is nothing we can do about this. If we do queries, we are still sysmodal and still have focus and have not changed our zorder (still on top) and have not been invalidated, yet there is progman on top of us. Pmwin needs to handle this scenario and any other that may occur because of this "forced paint". This is most easily recreatable by sticking a program object on the desktop with progman.exe as the executable and then starting it up, alt-esc'ing and then doing a lockup now off of the desktop context menu. After a few seconds, progman will appear, but cannot be used since lockup is still sysmodal. If you have any questions, please contact me, Aaron Reed at 2-XXXX or James Taylor. This is a tough problem. We are investigating possible fixes already under defect 59344. It is not clear that a fix will be possible as this affects many complex areas of the PM window Manager that may be too dangerous to change at this stage. Kelvin. 3-XXXX ═══ 2.280. 61822 BETA: RESTARTOBJECTS is not documented in the online cmdrefduplicate ═══ name 61822 reference abstract BETA: RESTARTOBJECTS is not documented in the online cmdrefduplicate state closed severity 3 answer docm symptom vis duplicate defects: none. duplicate features: none. notes: RESTARTOBJECTS is not documented in the online command reference. The SET command is used in many CONFIG.SYS statements to setup the Workplace shell environment. Documentation of all these statements is, given the Borg schedule, a subject for another release. This SET RESTARTOBJECTS= statement isn't a default statement and does not appear in the CONFIG.SYS so that it would need to be mentioned in the writeup of the SET AUTOSTART statement which is in the CONFIG.SYS file by default. Vince Spadea ID 2-XXXX I am going to include the RESTARTOBJECTS= statement in the list of Workplace Shell system variables in the existing section about system variables under the SET command documentation. Other Workplace CONFIG.SYS statements such as SET AUTOSTART= and RUNWORKPLACE= will also be included. Vince Spadea ID 2-XXXX Screen Team Approved ═══ 2.281. *61825 BETA: sys3176 error if no diskette in DRIVE A ═══ name 61825 reference abstract BETA: sys3176 error if no diskette in DRIVE A duplicate state closed severity 3 answer dsai symptom vis duplicate defects: none. duplicate features: none. notes: Whenever Drive A is accessed within a DOS session and NO diskette is present in the drive - the session gets a SYS3176 error. On driver 6.492 this problem is unreproducable. On 6.492 it is reproducable. IF a DOS FULL SCREEN is opened and DIR A: is done at the prompt - message SYS3176 IS displayed on drop 6.492. "DIR A:" with no diskette in the drive should generate a hard- error pop-up SYS0039 "The A: drive is not ready". Choices should be Abort(End Program), Retry, Fail (Return error code). Selecting Retry should do just that. If you have put a media in drive A: it should continue, if not another SYS0039 should pop-up. Selecting Abort should terminate the DIR command. Selecting Fail will return an error to COMMAND.COM. Unfortunately, all I get back from the hard-error pop-up is a message that the user decided to Fail the error. It doesn't tell me what error the user failed. Therefore the DIR command makes it's best guess. The guess turns out as SYS0021 "The drive is not ready". I have 6.492, plus the 6.494 additions on a machine in my office and I can not get the failure described in the defect. Ok... I've seen it now. This problem only occurs if DOS=HIGH. I'm investigating. Answer code is bogus.... Screen Team Approved ═══ 2.282. 61842 BETA: Refresh needed on dashed line in folder ═══ name 61842 reference abstract BETA: Refresh needed on dashed line in folder duplicate state canceled severity 4 answer intit symptom vis duplicate defects: none. duplicate features: none. notes: Open a folder with a few objects in it. Click on the left corner to get pull down menu. Notice the dashed line along the outside border of the folder. Press f1 for help. Notice that the dashed line is not contiguous any longer. Close the help folder and notice again that the dash line is no longer contiguous conti Screen Team Disapproved This problem has been noted before. It is related to the information we get regarding invalidated regions and how we repaint based on that. We would like to trs for now as it is not a borg ship issue. -Sheila 2-XXXX ═══ 2.283. 61859 clipboard does not replay metafiles properly ═══ name 61859 reference abstract clipboard does not replay metafiles properly duplicate state canceled severity 3 answer invextsw symptom vis duplicate defects: none. duplicate features: none. notes: The clipboard is not replaying metafiles copied from PMEXCEL correctly. Copy a part of a metafile to the clipboard and you will notice some of the vertical lines do not appear. Render it as a bitmap and all appears normal. Screen Team Approved I have investigated this from the clipboard viewer point of view We basically just call gpiplaymetafile to display the metafile. I need someone to look at this from the Graphics Engine point of view. Roberta 3-XXXX This defect was modified from component grensub per advise of Jimmy G. by K.Hamill on 2/18/93 I have done an addition test to ensure that this is not a clipboard viewer problem. I saved the excel clip to a file and viewed it with Picture Viewer ... it behaves the same way the vertical lines are not all appearing . The excel metafile can be provided ... Please contact Roberta Carey with an questions 3-XXXX. Brian, as we discussed today, I believe the problem is with Excel. The problem occurs when Excel generates the metafile. I've traced it through and found that for each box it draws it issues a GpiMove and and GpiBox command to draw the boxes. The first row of boxes (the ones containing the letters) are generated correctly. When they do the second row (the row(s) which would contain data, they only attempt to draw usually about 10 boxes, no matter how many have been selected. None of their calls fail; it's as if they only attempt to draw a certain number and then quit. I've tried to unassemble their code to understand what they are checking to determine how many boxes to draw (the missing lines are actually boxes) , but to no avail. I did find that they set the page viewport to a retangle large enough to encompass all the boxes. Finally I went back and tried this in 2.0. Although you can not render metafiles in the clipboard in 2.0, you can transfer data with it. If you generate this same metafile with Excel and paste it into Picview, you will see the same problem. This further leads me to believe the problem is with Excel. Debbi Best 3-XXXX ═══ 2.284. 61866 EAUTIL Produces wrong return code. ═══ name 61866 reference abstract EAUTIL Produces wrong return code. duplicate state open severity 4 answer symptom use duplicate defects: none. duplicate features: none. notes: If I use EAUTIL to split of EAs from a file to a hold file and the source file does not have any EAs, EAUTIL still ends with RC=0 even thought he hold file is not created. Should exit with RC>0. Screen Team Disapproved ═══ 2.285. 61904 BETA: Using PUBPB w/o VM option, focus fixed after session duplicate ═══ name 61904 reference abstract BETA: Using PUBPB w/o VM option, focus fixed after session duplicate state closed severity 3 answer chki symptom thipe duplicate defects: none. duplicate features: none. notes: If you use Publishers Paintbrush, OS/2 or Windows version without the Virtual Memory option for larger graphic files, the app woks fine. Close app and the focus cannot be moved with pointer and NO context menu for the desktop can be accessed. ANY icon you dbl-clk on opens the fixed focus icon instead. The keyboard allows movement of the focus and eventual shutdown using Shift-F10. after reboot focus fine using mouse. Test bed available w/DB terminal in 13-W240. The contact in the BTIG for this problem is Steve Rosbury. It looks like pmwin is getting into some state where WM_BUTTON1DOWN and WM_BUTTON1UP are not being translated into WM_CONTEXTMENU and WM_SELECT. Probably a problem in WinTranslateMouseMsg, but we did not go verify that. We just saw that the translated messages are never making it to the container. James Taylor Screen Team Approved Marty, Please can you look at this defect. Thanks. Kelvin First, the mouse click messages coming from this app and PMWORD (both are willow apps) are not synchronized. XScanSysQueue interprets the first two clicks as a double click message and set scanMouseDown=2. However, before the double click message reaches ProcessSqmsg, the scanMouseDown was reset back to 1 by a mousebutton1up message in xscansysqueue. Anyway, ProcessSQMSG should not need to check codeMousedown if it is processing a double click message. ═══ 2.286. 61932 BETA:PMDIARY.$$A file is missing from product ═══ name 61932 reference abstract BETA:PMDIARY.$$A file is missing from product duplicate state closed severity 3 answer bldi symptom codb duplicate defects: none. duplicate features: none. notes: BETA: User reported that we are no longer shipping pmdiary.$$a file in the product. This was shipped in previous versions. It is the tune file for the PMDDIARY application that comes in OS2\APPS directory. I checked with Gregg Mason and he found that it had been listed in the winos2 section of the build machine, but disklay.hst shows that on 10/02/92, Ayo Anise requested this along with several other files be removed. Ayo now tells me he knows nothing about that file and I believe he meant it shouldn't be listed in the winos2 section and not removed completely. It looks like we may have inadvertantly dropped a file that we were shipping in previous versions. Gregg asked me to write a defect to get the matter officially resolved. Screen Team Approved Arthur Roberts and I have determined that PMDIARY.$$A is necessary for the alarm clock and tune editor to work correctly. The file belongs in the \OS2\APPS directory. To test: (1) Open alarm clock from Productivity folder (2) Open "Alarms" menu and choose "Set alarm". (3) When the "Set Alarms" dialog appears, examine the "Tune" listbox at the right hand side. If the file is missing you will see something like: 0: Default Tune 49: Not in Use 50: Not in Use ... When the file is present, the list box will show a list of tunes, such as: 0: Default Tune 1: Blue Danube Waltz 2: Pop! Goes the Weasel For the tune editor: (1) Open it from Productivity folder. (2) Open "File" menu and choose "Open" (3) "File" listbox will show PMDIARY.$$A if it is present, otherwise the listbox will be empty. Therefore I have added PMDIARY.$$A back to the layout. Don Wakley 2-XXXX ═══ 2.287. 61962 Makeini not working correctly on the 492 driver ═══ name 61962 reference abstract Makeini not working correctly on the 492 driver duplicate state canceled severity 3 answer invtcp symptom use duplicate defects: none. duplicate features: none. notes: I was attempting to perform the 1.3 and WIN look alike desktops. I inserted the install diskette, then diskette 1, hit ESC to cancel, cd to d:\os2, erase os2.ini, makeini os2.ini os2_13.rc, received message that stated successful completion, then C-A-D. However upon rebooting the system hangs. IT can also be repeated when using the WIN_30.rc and OS2_20.rc. If you need any assistance, please contact Sheila McDonald-Rich on 3-XXXX. I contacted Val Robinson prior to opening. Thanks, I believe your test scenario is invalid. The look and feel .rc files are not intended to generate a complete .ini file. These are update .rc files and should be used against an existing os2.ini. Eliminate the step in your test where you delete the os2.ini and try it again. Lew Boehm (443-XXXX) ═══ 2.288. 61983 Gateway 486/66 system w/ATI Ultra Pro and Crystal Scan traps ═══ name 61983 reference abstract Gateway 486/66 system w/ATI Ultra Pro and Crystal Scan traps duplicate state canceled severity 3 answer invinf symptom thipe duplicate defects: none. duplicate features: none. notes: User getting traps and hangs using ATI Ultra Graphics Pro with a Crystal Scan 1572 monitor on a Gateway 2000 486/66. He had no problems with GA or 2.00.1 Beta. He has tried installing on HPFS and FAT. It appears to be related to display functions. He has faxed the trap information and the CONFIG.SYS. Need more information on this defect. Did his system trap during install, if so what diskette? Did it trap during reboot? A scenario, and more specific information would be helpful. Thanks. Screen Team Approved Customer found that his problem occurred mostly when he did a wildcard delete. Turned out he had the problem described in defect 57983. This defect can be cancelled ═══ 2.289. 61985 BETA: duplicate entries in settings pulldown menu for templduplicate ═══ name 61985 reference abstract BETA: duplicate entries in settings pulldown menu for templduplicate state canceled severity 4 answer chkmt symptom vis duplicate defects: none. duplicate features: none. notes: Use MB2 on a number of templates in the template folder, pick 'open' and get duplicate choices. For PIF file.PIF and Metafile.MET PICVIEW.EXE and Picture viewer are shown. For Bitmap.BMP and Pointer.PTR and ICON.ICO ICONEDIT, and Icon Editor are are shown. This is a duplicate of trs-ed defect 15314. We cannot dup against a defect in the returned state, so we're trs-ing this as well. It is not a borg ship issue. -Sheila 2-XXXX ═══ 2.290. 61988 BETA: Help for SYSTEM in SYSTEM folder incorrect ═══ name 61988 reference abstract BETA: Help for SYSTEM in SYSTEM folder incorrect duplicate state canceled severity 4 answer chkmt symptom vis duplicate defects: none. duplicate features: none. notes: In system setup folder, click once (highlight) the SYSTEM icon. Press F1 to get help. Last bullet says " Screen resolution for your display". The SCREEN page now ONLY appears if XGA is found. Rerouting from WPSH. The problem is that in the settings pages for the system object, the SCREEN page only shows up on machines with displays that support different resolutions. The help for the system object, however, just says that if you open me, I'll show you your screen resolution. To be more accurate it should say something like.... I'll show you your screen resolution if supported..... -Sheila 2-XXXX This would not appear to be a ship issue since a user can only viow the screen tab with certain display adapters present and when they are not present, the screen tab does not display and is thus of no further concern. Vince Spadea ID 2-XXXX ═══ 2.291. 62037 BETA: IN e.exe the FONT size of 24 pt is corrupted ═══ name 62037 reference abstract BETA: IN e.exe the FONT size of 24 pt is corrupted duplicate state canceled severity 4 answer asnit symptom vis duplicate defects: none. duplicate features: none. notes: Invoke E.EXE Select Options, Select Font size, Select Courier, Select Size There is a happy face next to 24 in the menu... This is a dup of 57793 and 36407 (among others) - which was TRSd. Screen Team Disapproved ═══ 2.292. 62041 BETA: fdisk error during install process using CD_ROM ═══ name 62041 reference abstract BETA: fdisk error during install process using CD_ROM duplicate state closed severity 3 answer chkm symptom vis duplicate defects: none. duplicate features: none. notes: model 90 - 2 harddrives - 1 CD-ROM drive - 1 SCSI adapter. Started loading OS/2 beta via CD-ROM. When asked if I wanted to accept the default partition setup I said no and went into the fdisk process. I defined the first disk as having boot manager and two other partitions and then went to the second disk where I deleted the the boot manager partition which was on that disk. I then deleted the other partitions on that disk and created just one partition. I then had to leave that drive and reenter to set the partition as bootable. I then selected the option to set start up values. FDisk failed at that point and brought me back to the initial menu. I then went back into FDisk and repeated all but the last step. When the system rebooted it errored. The install log indicated that fdisk had failed. System would not install after that. Screen Team Approved Jamie Jamison is working this defect -jamie is working this defect Added code to check that the boot manager is valid before using it and enabling its multi-boot functions. This will alert the user that a boot manager is not present, allowing the creation of one. ═══ 2.293. 62050 BETA: WIN_30.RC Observations ═══ name 62050 reference abstract BETA: WIN_30.RC Observations duplicate state returned severity 3 answer invinf symptom use duplicate defects: none. duplicate features: none. notes: Subject: WIN_30.RC Observations After making an OS/2 1.3 "look and feel" version of the OS/2 2.1 desktop, I ran "MAKEINI OS2.INI WIN_30.RC" to see what I could see. You might reference my "OS2_13.RC Observ o This point is a serious problem -- I opened my normal 2.1 desktop from the Windows-flavor (double clicked the icon). Then: - Opened a folder of miscellaneous objects - Started the FORBROWS application, but did not select a file - Started PMHPO (from the same folder) and dialed IBMLink - Attempted to switch to the FORBROWS application - EVERY attempt to get to FORBROWS actually got my to PMHPO - Eventually logged off IBMLink and closed the PMHPO application, which ALSO closed the FORBROWS application - During this, both icons in the folder were crosshatched until I shut down PMHPO, when both CLOSED If you need any assistance, please contact Sheila McDonald-Rich on 3-XXXX Thanks Is this problem unique to the win_30.rc scenario? In trying to switch apps from PMHPO, what method was used? We do not have the apps mentioned in the defect. Can this be reproduced on a machine with a debugger on it so we can take a look? Thanks -Sheila 2-XXXX ═══ 2.294. 62068 The following Windows 3.1 program behaves differently when star ═══ name 62068 reference abstract The following Windows 3.1 program behaves differently when star duplicate state canceled severity 3 answer invunr symptom fnwad duplicate defects: none. duplicate features: none. notes: The following Windows 3.1 program behaves differently when started seamlessly than when started from a WIN OS/2 full screen. The comments in the text explain. I have a compiled version if you want to see it. R. Viscuso 2-XXXX /* A problem occurs when running a windows application seamlessly from the OS/2 desktop. If the application consists of a parent window with several non-MDI child windows, calling BringWindowToTop() on one of these child windows does not bring the window to the top. Also, when a sibling window is created, it is created behind its other siblings. This behaviour is different from the behaviour observed under Windows 3.1. Note that if you start a WIN-OS2 full screen session and invoke the application from the WIN-OS2 session, the application runs correctly. The following is a sample that demonstrates this problem. When the program is running, use the left or right mouse button to select a child window and observe the behaviour described above. Again, note that the problem occurs when the session type is as a WIN-OS/2 window running as a separate session. */ #include typedef unsigned int UINT; #define STYLE WS_CAPTION | WS_OVERLAPPED | WS_THICKFRAME | WS_CLIPCHILDREN #define CHILD WS_CHILD | WS_CLIPSIBLINGS static int NumWindows = 0; long FAR PASCAL _loadds WindowProc( HWND hwnd, UINT msg, UINT wparam, LONG lpara { switch( msg ) { case WM_CREATE : NumWindows++; break; case WM_LBUTTONDOWN : case WM_RBUTTONDOWN : BringWindowToTop( hwnd ); break; case WM_DESTROY : NumWindows--; if( NumWindows == 0 ) { PostQuitMessage( 0 ); } break; } return( DefWindowProc( hwnd, msg, wparam, lparam ) ); } /* * WinMain */ int PASCAL WinMain( HANDLE hInstance, HANDLE hPrevInstance, LPSTR lpCmdLine, int nShowCmd ) { MSG msg; WNDCLASS wc; BOOL success; HWND parent; HWND child1; HWND child2; RECT rect; hPrevInstance = hPrevInstance; lpCmdLine = lpCmdLine; nShowCmd = nShowCmd; if( !hPrevInstance ) { wc.style = CS_DBLCLKS; wc.lpfnWndProc = WindowProc; wc.cbClsExtra = 0; wc.cbWndExtra = 0; wc.hInstance = hInstance; wc.hIcon = LoadIcon( NULL, IDI_APPLICATION ); wc.hCursor = LoadCursor( (HANDLE) NULL, IDC_ARROW ); wc.hbrBackground = GetStockObject( WHITE_BRUSH ); wc.lpszMenuName = NULL; wc.lpszClassName = "TestClass"; success = RegisterClass( &wc ); } if( success ) { NumWindows = 0; parent = CreateWindow( "TestClass", "Parent", STYLE | WS_SYSMENU, CW_USEDEFAULT, CW_USEDEFAULT, CW_USEDEFAULT, CW_USEDEFAULT, NULL, NULL, hInstance, 0 ); if( parent != NULL ) { ShowWindow( parent, SW_SHOW ); UpdateWindow( parent ); GetClientRect( parent, &rect ); child1 = CreateWindow( "TestClass", "Child1", STYLE | CHILD, 0, 0, rect.right, rect.bottom*3/4, parent, NULL, hInstance, 0 ); if( child1 != NULL ) { ShowWindow( child1, SW_SHOW ); UpdateWindow( child1 ); } child2 = CreateWindow( "TestClass", "Child2", STYLE | CHILD, 0, rect.bottom/4, rect.right, rect.bottom, parent, NULL, hInstance, 0 ); if( child2 != NULL ) { ShowWindow( child2, SW_SHOW ); UpdateWindow( child2 ); } } } if( NumWindows == 0 ) { PostQuitMessage( 0 ); } while( GetMessage( &msg, NULL, 0, 0 ) ) { TranslateMessage( &msg ); DispatchMessage( &msg ); } return( msg.wParam ); }  This is a duplicate of 58655, which was fixed in the driver immediately following the BETA. Mark Rogalski ═══ 2.295. *62079 BETA:SysCreateObject setup string not being handled correctduplicate ═══ name 62079 reference abstract BETA:SysCreateObject setup string not being handled correctduplicate state closed severity 3 answer chki symptom fnwad duplicate defects: duplicate features: none. notes: BETA: User submitted following REXX file which doesn't seem to work as expected. Object is created with MINIMIZED=YES and opens minimized. It is then closed and a SysSetObjectData is issued. Then when the object is opened, it is not opened minimized. Jeff Gray in Endicott was sent the file and he has indicated to me that he thinks the setup string specified on the call to SysCreateObject is not being handled correctly. /* */ Call RxFuncAdd 'SysLoadFuncs', 'RexxUtil', 'SysLoadFuncs' Call SysLoadFuncs Call SysCreateObject 'WPProgram', 'Test', '',, 'EXENAME=c:\OS2\CMD.EXE;MINIMIZED=YES;OBJECTID=;', 'R' Say 'Now open the object, it is opened minimized. Close it when done.' 'PAUSE' Call SysSetObjectData '', 'NOMOVE=YES;' Say 'Now open the object, it is NOT opened minimized, but it should...' Exit 0 Screen Team Approved tested in r205/6.503 and r206/6.504 is ok! ═══ 2.296. 62091 BETA: the help facility in QBASIC/EDITOR not found ═══ name 62091 reference abstract BETA: the help facility in QBASIC/EDITOR not found duplicate state open severity 4 answer symptom vis duplicate defects: none. duplicate features: none. notes: EDIT.HLP is missing. Asking for help within the qbasic/editor causes a message to display the EDIT.HLP can not be located. ═══ 2.297. 62095 BETA:Crosshatched Seamless Window Apps not really running ═══ name 62095 reference abstract BETA:Crosshatched Seamless Window Apps not really running duplicate state verify severity 3 answer inti symptom fnwad duplicate defects: duplicate features: none. notes: BETA: Numerous beta users have reported incidents where they have cross-hatched Window applications but they don't see them running nor are they in the Window List. I can now reproduce this problem consistantly. Drag a program template to the startup folder and give it a path to the clock.exe window application. On session page, indicate separate session. Shutdown and boot. The clock is started ok. Now drag a program icon to desktop and give it the path to notepad.exe window application. On session page , indicate separate session. Start the notepad. Shutdown with the clock and notepad active and then boot. You should get a System Error Program popup with many black holes where information should be. Sometimes this message is about the notepad app and sometimes about the clock. The message says- The process cannot access the file because it is being used by another process. If I click on okay button. I will have cross hatching on both apps, but only one is started and in the Window List. Looks like a timing problem somewhere. Running on mod 95 with VGA attached. I am not sure this is a WPSH problem. Please transfer it to appropriate component if I have picked the wrong one. Workplace is just starting the applications in a sequence that causes a file access clash. There shouldn't be a difference what order or how fast we start windows apps in separate sessions. I think you guys should look at this one first. I don't see how this is a workplace problem. --Mindy Screen Team Approved These defects are being reassigned to Ivan Lomelli. As per his request, this defect was reassigned to Carlos Romero. DOS Compatibility MODE opens are not allowed now at startup time. CMVC:Fixed incorrect AFLD2 field. CMVC: cloned to new defect 63620. ═══ 2.298. 62105 REGR: Text of Embedded Nulls Dialog incorrectly displayed. ═══ name 62105 reference abstract REGR: Text of Embedded Nulls Dialog incorrectly displayed. duplicate state verify severity 2 answer dsai symptom vis duplicate defects: duplicate features: none. notes: Severity based on APAR severity. PROBLEM: The text in the System Editor (E.EXE) "Embedded Nulls" Dialog is being truncated when using XR06055 and VGA resolution. This has also been observed in 6.492 Borg Release. To reproduce, try to edit a .EXE or .COM file. DIAGNOSIS: Examination of the dialog shows that the word "transformed" would have just exceeded the width of the control had it not wrapped to the second line. Since this is a relatively long word, the second line must now wrap to a third line and displays as a series of 4 dots since the third line is not fully displayed. I have checked the calculations used to create the dialog box and the text control and cannot find any obvious problems. I also verified that the value contained in cxSysFontChar matches that returned by GpiQueryFontMetrics for the default font -- no problem here. Using the Font Editor, I was able to determine that a few of the characters have apparently changed in width with the Service Pack, for example, "i" went from 4 pixels to 5 pixels, while "m" went from 12 pixels down to 11 pixels. Apparently these changes were not sufficient to change the average character width of the font, but do tend to affect the point at which a given string is forced to wrap. SUGGESTED RESOLUTION: In order to avoid the problem occuring in the future, it would probably be best to extend the dialog to handle a third line for this control. An alternative would be to increase the width of the control by 1 or 2 dialog units, but further modifications to the font might well cause the problem to recur. The above assumes that the character width changes were necessary for standards compliance. Should this not be true, then this defect should be rerouted to the font component to get the font corrected (since this will probably also hit our customer's applications as well). Screen team Approved Will give more room in the static control field to handle variations in font sizes. ═══ 2.299. 62118 BETA: EA size differs in dir /n and 2nd page notebook file ═══ name 62118 reference abstract BETA: EA size differs in dir /n and 2nd page notebook file duplicate state open severity 3 answer symptom vis duplicate defects: none. duplicate features: none. notes: Use config.sys. DIR config.sys /n to get hpfs format of the file. The EA will show 35. Open drives folder and get to config.sys icon and open settings. Go to 2nd page of FILE tab. It shows 156 bytes. Screen Team Approved ═══ 2.300. 62126 dual boot install - shell = c:\dos\command.com ═══ name 62126 reference abstract dual boot install - shell = c:\dos\command.com duplicate state closed severity 3 answer chki symptom comp duplicate defects: none. duplicate features: none. notes: If one attempts installation for Dual Boot feature, with the shell statement in config.sys contai sign install will fail. This should be addressed as the statement shell = c:\dos\command.com is perfectly legal in DOS's config.sys yet terminates installation of OS/2 Screen Team Approved Screen Team Approved Added appropriate parsing to allow for white space. -- Duane Wood ═══ 2.301. 62131 BETA: WINOS2 File Mgr does not preserve EA's copying OS/2 fduplicate ═══ name 62131 reference abstract BETA: WINOS2 File Mgr does not preserve EA's copying OS/2 fduplicate state working severity 2 answer intm symptom comp duplicate defects: none. duplicate features: none. notes: If you start WINOS2 and then star File Manager, do a copy of any file which has EA's and then do a ATTRIB /N on the file and it has no EA's listed. Actual Windows does not support EA's. This would require significant code addition to the file manager to be implemented. This should be considered for the next release. Mark Rogalski If a user copies his config.sys or any other OS/2 system file and deletes the original. The copy made by WINOS2 File Mgr would crash the OS, due to lack of EA's This Defect reopened per Irv Spalten . Screen Team Approved Added support to os2krnl, doskrnl, and VDPX.SYS so that DOS/DPMI applications can transfer EAs from one file to another. Added a worker so that DOS or DPMI applications can copy EAs from one file to another. ═══ 2.302. 62141 Windows Control Panel Drop Down Help Panel Transparent on XGA ═══ name 62141 reference abstract Windows Control Panel Drop Down Help Panel Transparent on XGA duplicate state canceled severity 3 answer valunr symptom vis duplicate defects: none. duplicate features: none. notes: The following problem has also been verified by Howard SIms. Howard has already discussed this problem with Poh. I am opening this defect at the request of Howard Sims. Probelm 1: Running on a Mod 95 with XGA-1 support. Installed latest driver #494 this morning. Migrated all WIndows Apps including Progman. Started Progman in a WIN-OS2 seamless session, Opened up control panel. Selected HELP. Under Help selected a green highlighted item. To reproduce the problem, the highlighted item needs to have a pop down menu associated with it. It doesn't happen all the time but when it happens the pop down menu is translucent and you see both the text in the pop down as well as the text associated with the item prior to the pop down. Moving the mouse button will cause a screen to repaint. This was with Control Panel taking up the whole screen. Second Problem If you window the control panel and repeat the steps above you will end up with entire window portions (several inches square) on the desktop that are never repainted. Screen Team Approved This is unreproducible under current OS/2 build 6.498. ═══ 2.303. 62192 BETA:Desktop Menu Help needs more info on applications ═══ name 62192 reference abstract BETA:Desktop Menu Help needs more info on applications duplicate state canceled severity 4 answer docmt symptom use duplicate defects: none. duplicate features: none. notes: BETA:Following was submitted on OS2BETA forum. Concerns incomplete explanation of use of applications that are added to the Desktop menu. Other people on the forum have also had problems in this area. ----- OS2BETA FORUM appended at 14:35:35 on 93/02/09 GMT (by KENT at TOROLAB6) Subject: Desktop Menu Help does not provide all the appropriate info Axiom: Nothing is fool proof..., fools are to creative!! In an effort to minimize the number of icons cluttering my desktop, I added an item to my desktop menu which allows me to create and OS/2 window from the desktop pop-up menu. My initial attempt resulted in SYS1034 error messages... So I opened defect 59961. The response which I got (a valid and helpful response) was as follows: the desktop pop-up menu is not supposed to be used to start applications. It is supposed to be used (menu page) to start applications that get the desktop passed in as a parameter. that was the design for the menu page. This was supposed to be instead of using associations. The only work around to make this work is at the program object that you found to set the menu item with (the OS/2 window) put %% in the parameters field. That way when it gets fired up the desktop parameter will be suppressed. I called the developer who responded, to discuss this and it became clear that the help for this feature needs to be clarified to explain what types of program objects should be added to the desktop pop-up menu. Can one of the Boca information developers please update the appropriate help text to clarify this point? Kent Smith OS/2 Compiler Development Sorry the Using the Operating System is closed for development of new information. ═══ 2.304. 62201 SOM.H Breaks WATCOM Compiler ═══ name 62201 reference abstract SOM.H Breaks WATCOM Compiler duplicate state working severity 3 answer chki symptom use duplicate defects: none. duplicate features: none. notes: I have sopken to David Kenner about this. He is aware of it and I am openening this defect with his ok. ******************************************************************** A customer has reported a problem using the SOM header files with our C compiler (WATCOM C/386). We checked it out and found that the following lines in SOM.H cause a problem when compiled with our compiler. #ifdef _CL386 #define SOMLINK _syscall #else #define SOMLINK #ifdef __IBMC__ #pragma linkage(somAssert║║ , system) . . . #pragma linkage(somTD_SOMRealloc║ , system) #endif #endif As you can see, when this is compiled with WATCOM C (or any C compiler other than CL386 or C Set/2), SOMLINK becomes empty which causes a problem since the SOM library requires the "system" calling convention. Can I propose that these lines in SOM.H be changed to the following. #ifdef __IBMC__ #define SOMLINK #pragma linkage(somAssert║║ , system) . . . #pragma linkage(somTD_SOMRealloc║ , system) #else #define SOMLINK _syscall #endif This will work for C Set/2, CL386 and WATCOM C. For other C compilers that don't recognize the "_syscall" keyword, an error will probably be issued. However, I beleive that a compile-time error is more appropriate than generating code that will not work. Screen Team Approved SOM.H breaks WATCOM Compiler ═══ 2.305. 62225 BETA: no general page for spooler icon ═══ name 62225 reference abstract BETA: no general page for spooler icon duplicate state closed severity 3 answer intm symptom vis duplicate defects: none. duplicate features: none. notes: NO General tab found for the spooler icon. Thus the icon can not be edited nor the title changed via the usual method. Screen Team Disapproved Screen Team Approved ═══ 2.306. 62244 Desktop pulldown ... accelorator keys conflict ═══ name 62244 reference abstract Desktop pulldown ... accelorator keys conflict duplicate defect 24584 state canceled severity 3 answer duplicate symptom use duplicate defects: none. duplicate features: none. notes: With the desktop pulldown menu, the accelerator key for Create Shadow (s) and Sort ( +S ) both bring up the window to create a shodow. To reproduce: 1) Click with the right mouse button on the desktop. 2) type "s" - the create shadow window will appear 3) Close the window and again click with the right mouse button on the desktop 4) type "S" - again the window for Create Shadow appears rather than the sort operation This is the third duplicate we've received on this from btig. Please check through cmvc before submitting -Sheila 2-XXXX ═══ 2.307. 62246 no documentation for accelerator keys on pulldown menus ═══ name 62246 reference abstract no documentation for accelerator keys on pulldown menus duplicate state canceled severity 3 answer valunr symptom vis duplicate defects: none. duplicate features: none. notes: When one creates a new menu item for a pulldown window, there is no documentation in any of the help facilities for adding the function of an accelerator key for the new menu item by using the tidle (~). ═══ 2.308. 62247 changing name of pulldown menu selection does not refresh ═══ name 62247 reference abstract changing name of pulldown menu selection does not refresh duplicate state closed severity 3 answer asni symptom vis duplicate defects: none. duplicate features: none. notes: If one tries to change the name of an option on a pulldown menu, it does not refresh the menu page or the pulldown until after the settings notebook for the pulldown is closed. To reproduce: 1) Go to the menu page of the settings notebook for any object. I used the desktop. 2) select Create Another from the available Menus 3) give it the name ~prompt 4) select OK 5) highlight prompt 6) select settings 7) change the name to ~prompts 8) select OK the name in the menu page does not refresh until the settings notebook is closed Screen Team Approved ═══ 2.309. 62248 adding a new menu item always creates cascade arrow ═══ name 62248 reference abstract adding a new menu item always creates cascade arrow duplicate state canceled severity 3 answer invwad symptom vis duplicate defects: none. duplicate features: none. notes: If one creates a new primary pop-up menu item in the menu page of the settungs notebook for an object, there is not any type of selection for an uncascaded item. When one looks at the pulldown for the object it will always have an arrow to the right to cascade to another menu, regardless of whether there is another menu to select This is working as designed. You are using the menu page wrong. If you want to create what you call an "uncascaded item," don't use the create another button in the "available menus" part of the dialog page. That will give you a new menu. Use the create another button in the "actions on menu" part of the dialog. The top half of the dialog is for creating new menu items (menus are either cascaded or conditional-cascaded). If you simply want to add an action to the context menu, then make sure "primary popup menu" is highlighted in the top box and then choose "create another" in the actions for that menu (bottom half of page). -Sheila 2-XXXX ═══ 2.310. 62292 BETA:Image Adapter Rexx SETDISP causes SYS3175 in REXX.DLL ═══ name 62292 reference abstract BETA:Image Adapter Rexx SETDISP causes SYS3175 in REXX.DLL duplicate defect 60797 state canceled severity 3 answer duplicate symptom thipe duplicate defects: none. duplicate features: none. notes: BETA-User reported and I have reproduced SYS3175 access violation in REXX.DLL. This program has worked successfully until driver 6.494. This program is run to enable user to use the Image Adapter display. cs:eip=005b.169e8955 ss:esp=006b.000efaa8 I ran this on mod70 with VGA. /* SETDISP.CMD - program to setup the ini for an old 16-bit display driver */ /* Initialize some variables. */ key.1 = 'CURRENTDRIVER' key.2 = 'DEFAULTDRIVER' key.3 = 'DISPLAY' keydata.1 = 'DISPLAY'||'00'x keydata.2 = 'DISPLAY'||'00'x keydata.3 = 'DISPLAY'||'00'x max = 3 /* First ensure that the REXXUTIL functions are registered. */ Call RxFuncAdd 'SysLoadFuncs','RexxUtil','SysLoadFuncs' Call SysLoadFuncs /* Update INI */ error = 4 Do i=1 to max If 'ERROR:' = SysIni(,'PM_DISPLAYDRIVERS',key.i,keydata.i) Then Do Say 'Error setting key "'key.i'".' Exit error End /* Do */ End /* do */ Screen Team Approved ═══ 2.311. 62314 BETA: Copyright notice shows wrong year in syslog.exe ═══ name 62314 reference abstract BETA: Copyright notice shows wrong year in syslog.exe duplicate state working severity 3 answer doci symptom vis duplicate defects: none. duplicate features: none. notes: Copyright screen in syslog.exe shows 1981, 1989 but not 1992 Screen Team Approved File which contains the copyright for this utility is under the utilmri component. The copyright probably should be updated to 1991. This is the last year in which a DCR-related change was made to syslog. Here is the response I got from an IBM lawyer about how copyright years should be dealt with: From: W. J. Brown IBM CONFIDENTIAL PREPARED BY IBM ATTORNEY Subject: (I) The answer can be found in Corporate StandardC-S 0-XXXX-002. It states that Date 1(1981 in this case) is the year of first external delivery of the copyrighted module. Date 2(1989) is the year of the first external distribution of the latest substantial changes to the module. If any of these dates are incorrect you should change them. Jim Brown 443-XXXX Assistant Counsel Intellectual Property Law - Boca Raton 91K/234-2 Internal Zip 4318 This is to update copyright in the syslog.exe program In order to change the copyright for syslog.exe, the file that must be changed is mri/cmd/syslog/pmsys.dlg. That file is owned by this component, UTILMRI! There is nothing I can do about this! ═══ 2.312. 62317 BETA:Database applet pushbutton should say Enter not Selectduplicate ═══ name 62317 reference abstract BETA:Database applet pushbutton should say Enter not Selectduplicate state open severity 4 answer symptom vis duplicate defects: none. duplicate features: none. notes: BETA:User has reported and I have confirmed that the Database application in the productivity folder (PMMBASE.EXE) has a mislabeled pushbutton. Under the DIAL function, it says "Select required number to dial and press 'Enter' or press 'Cancel' to leave panel without dialing." The pushbutton at the bottom says 'Select' rather than 'Enter'. Screen Team Disapproved ═══ 2.313. 62331 BETA:PALASM DOS Application system called failed ═══ name 62331 reference abstract BETA:PALASM DOS Application system called failed duplicate state closed severity 3 answer chkm symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA:Following is reported for Mike Turner (2-XXXX). Lisa Primm was made aware of the problem by Mike and she requested a defect be written so the problem could be looked at. -------------------------------------------------------------------- Problem summary: PALASM, a DOS application, uses DOS EXEC to run a number of programs. On OS/2 GA+SP, this program works fine. When running on driver 6_494 the program fails with a "system call failed, probably out of memory" message. There is more memory available under 6_494 than under the OS/2 GA+SP system. Mike Turner Screen Team Approved Needed to zero extend di after call to f_SFFromHandle in demDelete. PALASM had problem like Quicken(60981) except I had to add code to make fully-qualified pathname and go through all open handles to check for the file and close them before deleting. ═══ 2.314. 62340 BETA: Command ref example for Mouse incorrect ═══ name 62340 reference abstract BETA: Command ref example for Mouse incorrect duplicate state closed severity 3 answer doci symptom vis duplicate defects: none. duplicate features: none. notes: In the Command Ref for example: If an IBm mouse is installed the following statements are placed in the config.sys. DEVICE=D:\MDOS\VMOUSE.SYS is incorrect. It should be DEVICE=C:\OS2\MDOS\VMOUSE.SYS Will be fixed for Borg. 2/22/93 Vince Spadea ID 2-XXXX Screen Team Approved ═══ 2.315. 62343 BETA: Command ref example for LIBPATH incorrect ═══ name 62343 reference abstract BETA: Command ref example for LIBPATH incorrect duplicate state closed severity 3 answer doci symptom vis duplicate defects: none. duplicate features: none. notes: In the Command Ref for LIBPATH it reads: the install program places this statement in the config.sys LIBPATH=C:\OS2\DLL;C:\OS2\MODS;C:\; what is missing is the period IT should read: LIBPATH=.;C:\OS2\DLL;C:\OS2\MDOS;C:\; The line is also incorrect in the examples. Will be fixed for Borg. Screen Team Approved ═══ 2.316. 62344 BETA: Command ref example for SYSLEVEL incorrect ═══ name 62344 reference abstract BETA: Command ref example for SYSLEVEL incorrect duplicate state closed severity 3 answer doci symptom vis duplicate defects: none. duplicate features: none. notes: In the Command Ref for SYSLEVEL reads: C:\OS2\INSTALL\SYSLEVEL.OS2 IBM OS\2 Base Operating System Standard Edition 2.10 Component ID 560109001 Current CSD level: XR00000 Prior CSD level: XR00000 . . . o Standard Edition 2.10, Component ID: In fact, this is what you get: C:\OS2\INSTALL\SYSLEVEL.OS2 IBM OS/2 Base Operating System Version 2.10 Component ID 562107701 Type 0 Current CSD level: XR02010 Prior CSD level: XR02010 "Standard Edition" should be changed to "Version" in two places, the "Component ID" is incorrect, the "CSD levels" might be updated. and no mention is made of "TYPE 0." SYSLEVEL info will be updated for Borg. Vince Spadea 2-XXXX Screen Team Approved ═══ 2.317. 62345 BETA: Command ref example for IFS incorrect ═══ name 62345 reference abstract BETA: Command ref example for IFS incorrect duplicate state closed severity 3 answer doci symptom vis duplicate defects: none. duplicate features: none. notes: Some of the parameters that are actually put into the config.sys if IFS is selected are omitted from the discussion of putting HPFS in the system. Parms such as /CACHE:256 /CRECL:4 /AUTOCHECK:C The only one found is /C:64 in the example. And that may be incorrect, because it is system dependent what value C:## will be. The HPFS (IFS) information is found in the IFS(HPFS) command panel which can be linked to from the IFS panel. I am using a specific example for the CONFIG.SYS statement for IFS by showing the default statement for a 6MB system. Vince Spadea ID 2-XXXX Screen Team Approved ═══ 2.318. 62359 regr - IPE in file selmgra.asm at power-on ═══ name 62359 reference abstract regr - IPE in file selmgra.asm at power-on duplicate state verify severity 3 answer dsai symptom fnwad duplicate defects: duplicate features: none. notes: IPE at ##0170:fff44db2 - 000f:4db2. Line 60018, in file 4035 @(#)selmgra.asm. _SELSelToLa. IPE occurs when attempting to power-on or reboot Family II Mod 57 with plain VGA (machine #S438 in Generic Hardware Lab). Dan Heacock has seen the problem on the debugger. -- Patrick Flynn; Generic Hardware Lab; 2-4080. Screen Team Approved Loading multiple copies of the same PDD was going down the wrong code path, causing the selector manager a lot of problems. Fixed. Paul. ═══ 2.319. 62484 GrAF for Windows causes trap 14 when refreshing dde links ═══ name 62484 reference abstract GrAF for Windows causes trap 14 when refreshing dde links duplicate state verify severity 3 answer asnm symptom thipe duplicate defects: none. duplicate features: none. notes: GrAF, an as of yet unrealeased Windows application by Texas Instruments, causes Trap 14 when refreshing DDE link with orgmaint.exe( a demo also provided by TI ). This only occours when a refresh is selected and the orgmaint window is hidden. Trap 14 (0EH) - Page Fault 0004, Not Present, Read Access, User Mode eax=00000001 ebx=00000000 ecx=01d30000 edx=10c10007 esi=00000000 edi=00000000 eip=10bf5625 esp=01832f4c ebp=01832f68 iopl=2 rf -- -- nv up ei pl nz na po nc cs=005b ss=0053 ds=0053 es=0053 fs=150b gs=0000 cr2=01d30005 cr3=001d1000 005b:10bf5625 8b4905 mov ecx,dword ptr [ecx+05] ds:01d30005=invalid near symbols are: 005b:10bf55fe pmdde:_TEXT:_DdePipeWrite + 27 005b:10bf570a _DdeTransmit - e5 Screen Team Approved A memory object was being inappropriately freed. This was caused by not zeroing out a field in the "ConvNode" after it was used. Subsequent use of the structure cause the memory object referenced by this field to be was a valid memory object in use by another structure. When this memory object was referenced a trap 14 occured. ═══ 2.320. 62519 BETA: IOPL=NO in config causes boot hang ═══ name 62519 reference abstract BETA: IOPL=NO in config causes boot hang duplicate state working severity 3 answer inti symptom thipe duplicate defects: none. duplicate features: none. notes: If you change the IOPL line in config.sys to be NO, to disallow I/O privilages to programs, after rebooting OS/2 hangs. After all symbols have linked it stopped on the following: 05f8:00000579 xga : _TEXT:getmodeinfo + 9 05f8:00000594 getPOSdata - 12 I have a debug terminal and have tried this on a Mod 70,80 & 95. This isn't wpsh. I'm returning so you can reroute. -Sheila 2-XXXX Screen Team Approved Reopen to change to XGA component Allen will handle this matter. Always act as if IOPL=YES was specified, ignore whatever is entered in CONFIG.SYS ═══ 2.321. 62529 BETA: Color Palette - Effect on Klondike ═══ name 62529 reference abstract BETA: Color Palette - Effect on Klondike duplicate state open severity 4 answer symptom use duplicate defects: none. duplicate features: none. notes: Per my conversation with Donald Hobern, I am opening this defect in the sol component. TEXT: I mentioned some time ago (not in this CFORUM) the effect a system-wide change had on the Solitaire-Klondike game, but let me repeat it here. Dragging any color, other than white (top right corner color of the color palette) with the "Alt" key depressed will radically change the Klondike game. The card border becomes RED, totally obliterating the Hearts and Diamonds cards Ace thru Ten. This happens even if the game is not running Once Klondike is started you cannot drag the white color to the game and fix it-- you must make a system-wide change to some folder to make the game playable. I've only tried this with the default SVGA 640X480X16 drivers. If any questions, please contact Sheila McDonald-Rich on 3-XXXX. Thanks Screen Team Disapproved ═══ 2.322. 62538 BETA: PMSEEK -- WIndows Not Adjustable ═══ name 62538 reference abstract BETA: PMSEEK -- WIndows Not Adjustable duplicate state canceled severity 4 answer invwad symptom use duplicate defects: none. duplicate features: none. notes: Subject: PMSEEK -- Windows Not Adjustable I noticed two things about the PMSEEK windows. First, the basic window cannot be sized on my VGA (640x480x16) display. Grabbing the yellow border with the cursor does no good, even though the cursor turns into double-headed arrows. Second is the window containing "hits." If the result is "just right" in that the last character of the filename lies on the edge of the window, you cannot scroll the window horizontally to see all of that last character. For example, I have several files named "E:\OS2BBS\OS2BETA.931" (and 932, 933, 934, etc.). If I set the search argument for "Drive E" and "*.93*" then the last digit is unreadable. And, like I said, the horizontal slider is inoperative. The window frame of PMSEEK is not sizable. Screen Team Disapproved ═══ 2.323. 62548 BETA: title MLE cursor is same color as bkrnd in notebook -duplicate ═══ name 62548 reference abstract BETA: title MLE cursor is same color as bkrnd in notebook -duplicate state closed severity 3 answer asni symptom vis duplicate defects: none. duplicate features: none. notes: Open the settings notebook for a program object and go to the general page. The cursor for the title multi line entry field is the same color as the MLE's backround making it difficult if not impossible to position the cursor. This problem did not exist in the dec/92 beta but now does. Rerouting from wpsh. Please call me if this isn't the correct component. I talked to the submitter, and this appears to only be a problem on SVGA. He has the TSENG driver (ET4000). -Sheila 2-XXXX Screen Team Approved ═══ 2.324. 62549 BETA: title MLE cursor is same color as bkrnd in notebook -duplicate ═══ name 62549 reference abstract BETA: title MLE cursor is same color as bkrnd in notebook -duplicate state canceled severity 3 answer symptom vis duplicate defects: none. duplicate features: none. notes: Open the settings notebook for a program object and go to the general page. The cursor for the title multi line entry field is the same color as the MLE's backround making it difficult if not impossible to position the cursor. This problem did not exist in the dec/92 beta but now does. Duplicate of 62548 - Accidently hit enter in PM CMVC screen instead of quit ═══ 2.325. 62550 BETA: Selective install of SVGA on TSENG ET4000 kills WIN/Oduplicate ═══ name 62550 reference abstract BETA: Selective install of SVGA on TSENG ET4000 kills WIN/Oduplicate state canceled severity 3 answer invunr symptom vis duplicate defects: none. duplicate features: none. notes: Installed 6.494 driver on machine with TSENG ET4000 chipset. Everything worked fine with plain vanilla VGA drivers and fonts. Used selective install to bring system to 800x600x256. Selective install ran with no errors, but upon rebooting received "Can't load atmsys.drv" when trying to run winos2.com from os/2 desktop icon or full screen dos session. In addition any time I would go into a full screen dos session the second time around I would get THREE images of my dos app running crunched up side by side along with a high pitched tone. The 6.490 driver definitely did not have these problems. Screen Team Approved Returning defect 62550 because this problem was fixed in 6.496. ═══ 2.326. 62551 BETA: Borland C++ was incorrectly migrated as an OS/2 app ═══ name 62551 reference abstract BETA: Borland C++ was incorrectly migrated as an OS/2 app duplicate state canceled severity 3 answer invunr symptom vis duplicate defects: none. duplicate features: none. notes: Borland C++ 3.1 migrates incorrectly as an OS/2 app Fixed after Beta. ═══ 2.327. 62554 BETA: Adaptek 1542B does not show up first time in sel. insduplicate ═══ name 62554 reference abstract BETA: Adaptek 1542B does not show up first time in sel. insduplicate state canceled severity 3 answer invinf symptom vis duplicate defects: none. duplicate features: none. notes: Customer went thru entire install O.K. Support for Adaptek 1542B was installed as well as Toshiba 3401 CD-ROM. The first time user went into selective install the Adaptek card was not listed. Screen Team Approved Please add whether the 154xpress.exe presence check failed (non-zero) or not to the defect's notes. ═══ 2.328. 62614 Access of Unformatted Diskette Can Cause Stack Overflow ═══ name 62614 reference abstract Access of Unformatted Diskette Can Cause Stack Overflow duplicate state canceled severity 3 answer intit symptom thipe duplicate defects: none. duplicate features: none. notes: This problem was reported on 2.1 BETA and exists on the 6.494 internal drop. Put an unformatted diskette into the "A" drive. Click on the Drive A Icon. An error message is generated "The drive in diskette is not formatted correctly." Clicking repeated on Drive A Icon will create a stack of these messages such that the user gets the impression that he is unable to close the dialog box. After the first error message the Drive A Icon should not access the drive again until the pending error message box is disposed of by user action. Further - after clicking on the Drive A Icon about 17 times the system desktop totally collapses and System Error -Commands Prompts SYS3171 appears. Program caused exception C0000005 at 1A028670 due to insufficient stack space, the exception was not dispatched. The process terminated. DOSCALL1.DLL 0002.00008670 After closing the dialog box on this error. The system thrashed for a while and recreated the desktop. For a moment I thought I had lost everything. We have looked into this problem. Changing WPS so that the dialog boxes put up by ReportError requires making them system modal, which is a significant code hit. We will consider for the next release. -Sheila 2-XXXX ═══ 2.329. 62652 VGA256C Driver on Thinkpad 700C ═══ name 62652 reference abstract VGA256C Driver on Thinkpad 700C duplicate state verify severity 3 answer lfm symptom vis duplicate defects: none. duplicate features: none. notes: I will provide a copy of the testcase on diskette to Eric Farr. TESTCASE: (BATTDATA.EXE). Following this note I am sending you a copy as BATTDATA EXEBIN. Just start it up from the OS/2 command prompt, move pointer over it, then press and hold the right mouse button. A crosshair should appear but as you drag it (still holding down the button), the dotted grids will appear to get erased near the intersection. The program is writing text and line with mixing mode FM_XOR. Somehow the line gets drawn and erased OK, but the text isn't (it's overpainting the grids). The same program displays moving text (coordinate position) and cross- hair on other display drivers, including Trident and 8514/A.and From: CELI --BCRVMPC1 Date and time 02/22/93 09:34:19 To: V$IBITG7--BCRVMPC1 From: Joe Celi Subject: NO SUBJECT Mike, Please write up a defect for this problem. Please add Ken's comments and a pointer to the testcase. Thanks, Joe Celi CELI at BCRVMPC1 OS/2 Development Dept 6J5 Boca Raton, Florida (407)-982-XXXX Internal Zip 1510 *** Forwarding note from V$IBITG7--BCRVMPC1 02/20/93 10:36 *** To: CELI --BCRVMPC1 Joe Celi From: Mike Jasinski Subject: NO SUBJECT Joe, Are you aware of this problem, did Ken get in touch with you? Mike BPC Technical Interface Group tl: 443-XXXX 407 443-XXXX 407-982-XXXX FAX *** Forwarding note from JL17356 --YMTVM1 02/09/93 11:00 *** ========================================================================= Date: 9 February 93, 09:23:23 JST From: Ken Inoue JL17356 at YMTVM1 System Design, LAB-R11 Portable Systems, Yamato To: V$IBITG7 at BCRVMPC1 Re: Forwarding note re: VGA256C driver on Thinkpad 700C Mike, I got 6.490 from Bob to verify fix of another problem and noticed that the XOR handling under 256 colors mode apparently hasn't changed. I was told to contact Joe Celi - should I or have you folks written off this one? Best regards, Ken Inoue ***** Note follows ***** Date: 25 January 93, 18:21:19 JST From: Ken Inoue JL17356 at YMTVM1 System Design, LAB-R11-A Yamato To: V$IBITG7 at BCRVMPC1 Ref: Append to OS2BETA FORUM (IBMPC) at 16:10:11 on 93/01/22 by V$IBITG7 at BCRVMPC1. Mike, The program is now on OS2TOOLS as a part of the T700BATT PACKAGE (BATTDATA.EXE). Following this note I am sending you a copy as BATTDATA EXEBIN. Just start it up from the OS/2 command prompt, move pointer over it, then press and hold the right mouse button. A crosshair should appear but as you drag it (still holding down the button), the dotted grids will appear to get erased near the intersection. The program is writing text and line with mixing mode FM_XOR. Somehow the line gets drawn and erased OK, but the text isn't (it's overpainting the grids). The same program displays moving text (coordinate position) and cross- hair on other display drivers, including Trident and 8514/A. Hope this helps. Ken Inoue ***** Append follows ***** Subject: VGA256C driver on Thinkpad 700C Ref: Append at 10:09:37 on 93/01/19 GMT (by JL17356 at YMTVM1) Ken, Please send a copy of your sample program to my reader, along with a note reference this append. Provide ant pertinent instructions necessary to recreate the failure, and we will look at it. Thanks, Mike Jasinski Boca Technical Interface Group Screen Team Approved Add font XOR support. ═══ 2.330. 62659 tree view of drive header needs : ═══ name 62659 reference abstract tree view of drive header needs : duplicate state canceled severity 3 answer invwad symptom vis duplicate defects: none. duplicate features: none. notes: the tree view of a drive folder needs to contain a : when the drive is displayed in the header. 1) Open up the drives folder 2) Open up the drive OS/2 is installed on in a tree view 3) select the + next to OS2 folder 4) Note the drive information in the header says X\OS2 when to be visually correct it should be X:\OS2 The status information displayed in the tree view of a drive object is to help the user identify what is currently selected in an open tree view. It is not designed to be the fully qualified path of that folder. The Drive C string in the window is the name of the disk object, which does not have a : at the end. Changing that would give inconsistent information to the user. The names of the folders displayed in tree view are long names, not the actual file names (which are different for example on FAT drives), so making the Drive c..... string look more like a path name would be misleading. Usability agrees with this design issue. -Sheila 2-XXXX ═══ 2.331. 62669 BETA: System Clock -- Change Font ═══ name 62669 reference abstract BETA: System Clock -- Change Font duplicate state canceled severity 3 answer intit symptom use duplicate defects: none. duplicate features: none. notes: Subject: System Clock -- Change font If you open the Settings for the System Clock, then go to page two of "view," you get the opportunity to "Change font." I'm not sure WHAT it is you can do here, but it surely isn't change the fonts for the date and time on the digital clock. The Help implies the expected: You can alter the appearance. When you select "Change font," what you get is a way to install or delete fonts. A BETA TESTER If you have any questions, please contact SHeila McDOnald-Rich on 3-XXXX. Thanks, This is a duplicate of 14283, which has been trs-ed already. We are currently going through our trs backlog to prioritize which to include to enhance the borg product on our current schedule. 14283 is included in this review, as will this defect. we will trs for now. -Sheila 2-XXXX ═══ 2.332. 62772 BETA:Second occurrence of PRINT02.SYS in config causes IPE ═══ name 62772 reference abstract BETA:Second occurrence of PRINT02.SYS in config causes IPE duplicate defect 62359 state canceled severity 3 answer duplicate symptom thipe duplicate defects: none. duplicate features: none. notes: BETA-User reported on 6.479 driver that when he accidentally appended his config.sys to itself and booted, his system trapped. In trying to reproduce this problem on 6.496, I get an IPE at boot whenever I have 2 occurences of BASEDEV=PRINT02.SYS in my config.sys. Internal Processing error- ##0160:fffbadb8-000d:00063db8 1152,7019 Please transfer this to the correct component if this is not the right one. Known problem. Paul. ═══ 2.333. 62794 BETA: Wordperfect - notebook tabs dont left-justify text ═══ name 62794 reference abstract BETA: Wordperfect - notebook tabs dont left-justify text duplicate defect 57928 state canceled severity 3 answer duplicate symptom fnwad duplicate defects: none. duplicate features: none. notes: Tab text in notebook tabs will accept BKS_TABTEXTRIGHT and BKS_TABTEXTCENTER, but when BKS_TABTEXTLEFT is selected the tab text remains in the center, which is the default. In other words the BKS_TABTEXTLEFT flag is ignored. I have a notebook sample that demonstrates this. Dan Lee 2-XXXX Screen Team Approved ═══ 2.334. 62820 The on-line CP Program Reference has an error. Click on Control ═══ name 62820 reference abstract The on-line CP Program Reference has an error. Click on Control duplicate state closed severity 3 answer doci symptom use duplicate defects: none. duplicate features: none. notes: The on-line CP Program Reference has an error. Click on Control Program Funcions, click on DosGetInfoBlocks,Click on Function Syntax. The defintions read: PTIB *pptib; PPIB *pppib; They should read: PTIB pptib; PPIB pppob; Reassigning to the correct component CPPROGREF. Vince Spadea ID 2-XXXX Red asked to have this returned for further invetigation. Dan Baker I think this may be technically correct but confusing. The DosGetInfoBlocks call really looks for the address of. The * accomplishes the same thing but ........... Screen Team Approved I think a better way would be: PTIB pptib; PPIB pppib; Then you would call with: DosGetInfoBlocks( &pptib, &pppib); This is more in line with how most programmers code and is also the way the samples do it. See for example, the WORMS.C file in the WORMS sample in the ToolKit. The following are the changes made in the doc. for this defect. #define INCL_DOSPROCESS #include PTIB ptib; /* Address of a pointer to a TIB */ PPIB ppib; /* Address of a pointer to a PIB */ APIRET rc; /* Return code */ rc = DosGetInfoBlocks(&ptib, &ppib); Dan Baker - DJBAKER at BCRVMPC1 Correction to earlier note. The following are the changes made in the doc. for this defect. #define INCL_DOSPROCESS #include PTIB pptib; /* Address of a pointer to a TIB */ PPIB pppib; /* Address of a pointer to a PIB */ APIRET rc; /* Return code */ rc = DosGetInfoBlocks(&ptib, &ppib); Dan Baker - DJBAKER at BCRVMPC1 ═══ 2.335. 62821 The CP Program Reference (S10G-6263) has an error on page A-20. ═══ name 62821 reference abstract The CP Program Reference (S10G-6263) has an error on page A-20. duplicate state canceled severity 3 answer invunr symptom use duplicate defects: none. duplicate features: none. notes: The CP Program Reference (S10G-XXXX) has an error on page A-20. The definition of the members of the TIB2 structure SHOULD read: typedef struct _TIB2{ ULONG tib2_ultid; ULONG tib2_ulpri; ULONG tib2_version; ULONG tib2_usMCCount; USHORT tib2_fMCForceFlag; } TIB2; Please reassign to the correct component under the tool kit. ID_books is the component for the base product books. Thank you, Anita Soto 2-XXXX. . I am unable to reassign it. System tells me it must be open. Can you reassign to cpprogref ? Thanks. This problem was fixed in the CP Ref. Online info - See latest build - click on DATATYPES - select TIB2. The datatype section of the HARDCOPY (BOOK) for CP Ref had a lot of problems with the datatypes not matching the headers. ID has fixed this problem. It can be verified by looking at the online CP Ref doc. in the latest build. The hardcopy will reflect these changes also when the next issue is produced. There is no plan at this time to produce the hardcopy. Dan Baker - 24591 - DJBAKER at BCRVMPC1 ═══ 2.336. 62859 BETA: SYS3175 when starting Calculator ═══ name 62859 reference abstract BETA: SYS3175 when starting Calculator duplicate state canceled severity 3 answer invinf symptom vis duplicate defects: none. duplicate features: none. notes: On a system with 8MB RAM, with 4MB swap file and 11MB free space on disk. 1. Create a folder on the desktop and open it. 2. From the migrated WIN/OS2 accessories, copy all objects exept Media Player to Folder (9 apps). 3. From the Productivity folder, copy 21 objects (no softerm, pmchart, or fax/pm apps) to folder. The OS/2 calculator became Calculator.1. 4. Sort Folder by Name. 5. Select all objects in folder. 6. Start them. When the Win/OS2 calculator starts it generates a SYS3175 error. Screen Team Approved SYS3175: A program in this session encountered a problem and cannot continue. EXPLANATION: An access violation exception occurred and was generated when an attempt was made either to load or store data in an inaccessible location, or to execute an inaccessible instruction. This exception corresponds to both the Intel 80386 processor general protection fault (#13), caused by an invalid access attempt, and the page fault (#14), caused by an attempt to access an uncommitted page or a page with incorrect attributes for the desired operation. ACTION: If you purchased this program, contact the supplier of the program. If you are the developer of this program, refer to the information in the register. On driver 498, I tried to migrate all Windows application from the system folder, and I also get this sys3175 error screen dialog pop up. So the problem is in the os/2 kernel. My suspection is in the swapper or memory management. The description "When the Win/OS2 calculator start", that means the calculator is already up and running. It could be that the OS/2 operation system was trying to load the next program in the list. Joe Luu 26843 ═══ 2.337. 62876 BETA: screen corruption particular to Western Digital chipsduplicate ═══ name 62876 reference abstract BETA: screen corruption particular to Western Digital chipsduplicate state closed severity 3 answer asnm symptom vis duplicate defects: none. duplicate features: none. notes: Scenerio I: started clock.exe from startup.cmd on Dell 486/25p in VGA mode. Clicked on Win/OS2 minimized icon to switch to clock running in a full screen Win/OS2 session. Dragged clock accross screen. Problems encountered painting screen. Switched back to OS/2 desktop and back to clock. Clock couldn't gain focus. Assorted other anomolies. Scenerio II: started clock.exe from a full screen OS/2 session on an AST machine with WD chipset. Switched back and forth between full screen clock and OS/2 desktop. Corruption in OS/2 desktop appears. Machine was in 1024 x 768 resolution. First scenario couldn't be reproduced on 6.498. Second scenario is not WD specific. The corruption is due to the fact that desktop dirty flag is not set on the switch to the desktop, so desktop is not repainting. Screen Team Approved Scenarion 1 couldn't be reproduced. Scenario 2 is fixed by having an instance flag which ensures that on vdm's first foreground, desktop dirty flag is set. ═══ 2.338. 62882 DMQS Files Not Handled Properly ═══ name 62882 reference abstract DMQS Files Not Handled Properly duplicate state open severity 4 answer symptom thipe duplicate defects: none. duplicate features: none. notes: Note: Howard Sims has verifed that this problem exits in the latest driver. Subject: XGA-2 DMQS Support To: V$IBITG7--BCRVMPC1 From: Joe Celi Subject: XGA-2 DMQS Support Mike, They are both potential bugs in the XGA PM driver. These type of things are simply not tested by our test groups. Unfortunately, I don't have the bandwidth to investigate these problems. If someone can reproduce these problems then we should create individual defects for them so that we don't forget about them. These problems should not be ship issues and written as SEV 4's. Also, we would definitely need Randi's DMQS files. Regards, Joe Celi CELI at BCRVMPC1 OS/2 Development Dept 6J5 Boca Raton, Florida (407)-982-XXXX Internal Zip 1510 *** Forwarding note from V$IBITG7--BCRVMPC1 02/19/93 13:17 *** To: CELI --BCRVMPC1 Joe Celi From: Mike Jasinski Subject: XGA-2 DMQS Support Joe, Can I please have your comments on the attached. Is this a problem, and is there a work around. Thanks, Mike ============================================================================== ----- OS2BETA FORUM appended at 17:02:53 on 93/02/17 GMT (by RANDOB at BCRVM5) Subject: XGA driver problem with certain resolutions I've been developing XGA-2 DMQS files, and I've found that certain resolutions such as 1152x900 and 848x1200 don't work with OS/2. This is because they almost fill the 1MB VRAM and the XGA driver needs more off- screen VRAM than is available. I would like to have the driver sense how much off-screen VRAM it needs, and then appropriately determine how many colors it can use. In the case of these two modes, the driver is trying to use 256 colors, when it really should only use 16. I will be happy to supply test DMQS files. The 848x1200 is an 8506 mode, which I am plan- ning to ship with V2.03 of the XGA DOS device drivers. Thanks, Randy Dobkin - EST Graphics Subsystems Development Subject: XGA driver resolution problem To: V$IBITG7--BCRVMPC1 Michael *CONTRACTO From: Randy Dobkin EST Graphics Subsystems Development Dept. KVJ, 443-XXXX Subject: XGA driver resolution problem Mike, first you need to get a multiscanning monitor such as a NEC 5FG or PS/ValuePoint 6314 or 6319. I sent you MONE560.DGS. Put it in your \xga$dmqs directory. Then go to page 2 of the System notebook and select "Eizo (Nanao) T560i" for the monitor. Then shut down and reboot. Go to the Screen tab of the System notebook and select 1152x900 resolution and shut down and reboot. When the system comes up, you will notice that the characters are back- wards and the system may hang. The driver needs to sense it doesn't have enough off-screen VRAM and only use 16 colors. Thanks, Randy P.S. To get your system up again you will need to boot from diskette and erase \xga$dmqs\xgasetup.pro. BPC Technical Interface Group tl: 443-XXXX 407 443-XXXX 407-982-XXXX FAX Screen Team Disapproved ═══ 2.339. 62895 'Flags' in drive settings are set back to default after rebooti ═══ name 62895 reference abstract 'Flags' in drive settings are set back to default after rebooti duplicate defect 46524 state canceled severity 3 answer duplicate symptom use duplicate defects: none. duplicate features: none. notes: Scenario: 1) Open the Settings Notebook for a drive icon. 2) Select the Include Tab and go to page 2 3) Add or change the current 'flags' 4) Reboot. After rebooting the added flags disappear and the changed ones go back to their default values. This is a dup of apar 46524. -Sheila 2-XXXX Wrote a duplicate...Cancelling. ═══ 2.340. 62897 Changing path using the 'Set' option in Selective Install fails ═══ name 62897 reference abstract Changing path using the 'Set' option in Selective Install fails duplicate state canceled severity 3 answer valunr symptom use duplicate defects: none. duplicate features: none. notes: When doing a Selective Install, the user is prompted to enter the drive where the new files are. The user is presented with a list of possible drives. If the user selects a drive by double-clicking on the desired drive letter, then this works. If the user selects a drive and then hits the 'SET' option, then this fails. Screen Team Approved Neither myself or Carmen could reproduce the problem on the 6.498 driver. The dosOpen for a drive if the drive was network drive failed, with a return code of 65 (ERROR_NETWORK_ACCESS_DENIED). Changed the code to open the file [drive]:\[sourcepath]\disk_1\disk.num instead of doing an open on the drive. Please disregard the accept and corresponding note regarding the DOSOPEN this should have been for defect 62970 ═══ 2.341. 62913 BETA: dropdown list functions improperly under win/os2 ═══ name 62913 reference abstract BETA: dropdown list functions improperly under win/os2 duplicate state closed severity 3 answer chki symptom vis duplicate defects: none. duplicate features: none. notes: Start program manager in a seamless Win/OS2 session. Open a window, such as OS/2 Main. Go to the file menu, and select New. Select new program icon. When you get the window asking for the icon's information, click on the browse button. When the browse window comes up, click on the drive list. You can click on either the drop-down button or the display area, as either will drop down the list. Once the window is dropped down, listing the available drives, click on either the drop down button, or the display area. Instead of closing the window, and leaving the selected drive the same, your system should beeb, and change the Win/OS2 cursor to the OS/2 black cursor. To close the window, you should have to select a drive from the list. If you go over to the other drop down window, the Program Type window, and do the same thing, it will work properly. This problem apparently doesn't occur with every drop-down window, but is consistent with those that don't work. This happens only in seamless Win/OS2 sessions. For more information contact Richard Decker BTIG at x22457. When this scenario is followed, I noticed that Windows is receiving mouse event messages no matter where the mouse cursor is on the screen, and at the same time PM is drawing the black PM cursor everywhere on the screen except for the seamless list box. This tells me that something is wrong with the way PM win is working, because PM should never display the mouse cursor AND send a mouse event to Windows. Tachi and I both looked at this and suspect PMWIN. Please call if you have any questions. Mark Rogalski 2-XXXX Screen Team Approved I am routing this defect back to the seamless component. The drop down window is being created by seamless as a system modal window and therefore all mouse hittesting is getting warped and the mouse pointer is getting redrawn by seamless over the top of the PM pointer by the seamlessmouse draw code. I have spoken to the seamless folks about this. To see more clearly put a break register on pwndsysmodal using .ss8 ; br w2 pwndsysmodal. Run program manager seamlessly and got to the file dialog etc. When the break register hits ln will show that you are in winSetSysModalWindow from seamlesscreate. Hope this helps, Kelvin. 3-XXXX The problem is an incorrect define for WINWS _SYSMODAL in seamless.h. Windows.h defines DS_SYSMODAL and SDS_SYSMODAL as 0x0002, however, these are not defines for the STYLE field. I believe that MB_SYSTEMMODAL, defined as 0x1000, describes the bit we are interested in. Mark Rogalski ═══ 2.342. 62946 BETA: PMChess gets SYS3175 on exit using LS machine. ═══ name 62946 reference abstract BETA: PMChess gets SYS3175 on exit using LS machine. duplicate state canceled severity 3 answer invunr symptom thipe duplicate defects: none. duplicate features: none. notes: Running 6.479c on machine with LS functional, run PMChess app, close app and get SYS3175 on MAILSLOT.DLL debug LN shows following: 0170:fff4810 os2krnl:DOSHIGH32CODE:_tkSchedNext + 53 0170:fff4818 HPTDispatch -25 After trying again I Trapped with following: Trap 14(0EH) - Page fault 0004, Not Present, Read Access, User Mode ab6f:00001352 26803d01 cmp byte ptr es:[di],01 Debug available in System Test Lab "A", call 2-XXXX. Screen Team Approved Steve has verified that this is not reproducable on the 6.498 Beta level driver. This Defect is CANC due to not reproducable error on 6.500 with LS installed. SesMgr looks fine... ═══ 2.343. 62987 BETA: No CDROM disk change ═══ name 62987 reference abstract BETA: No CDROM disk change duplicate state working severity 3 answer dsai symptom fnwad duplicate defects: none. duplicate features: none. notes: This program also fails in OS/2 2.0 GA but works in OS/2 1.3. CDROMCHK is a program that call MSCDEX (when running under DOS) to read a cdrom. The program works just fine the first time it is run a VDM but if you change the cdrom, it continues to report the information from the first cdrom until you close the VDM and reopen it. I have verified that this is a bug still existent on the latest Beta Driver. The problem is thought to be in the MSDEX emulation in OS/2 2.X. I have spoken with Frank Schroeder, gave him the testcase and reproduced the problem for him. Thanks, Sheila McDOnald-Rich Screen Team Approved The problem is that the MSCDEX api and the DOS device driver are not reopenning the drive handle on each request. If the drive is changed between an open and a close request, vcdrom is not aware of it. The handle must be closed and reopenned so that the file system will remount the volume. There are at least two solutions to this problem. First, vcdrom must install a PDBDestroy user hook so that when an application terminates, the drive handles will be closed. It is very likely that a CD-ROM disc change will occur inbetween running two CD-ROM applications. The second solution is to open and close the drive handle before and after each VDHFSCtl call to guarrantee that the drive is valid for the VDHFSCtl call. This is costly in terms of performance. Therefore, I have added a new FSCtl call into the CD-ROM File System to do a media check on the drive. If there is an existing file handle, a media check is performed and if the media has changed, the media is reset. A reset is done by closing and then openning the drive. By doing a media check instead of open/close, performance is improved and the close/ open penalty only occurs when the disc actually changes. ═══ 2.344. 62990 DevIOCtl Cat4 Fun 56h Not WAD ═══ name 62990 reference abstract DevIOCtl Cat4 Fun 56h Not WAD duplicate state canceled severity 3 answer invunr symptom fnwad duplicate defects: none. duplicate features: none. notes: The attached test case fails to woak as documented. The DevIOCtl a Category 4, function 56H and a HotKey ID of -1 is supposed to disable Ctrl+Alt+Del, Ctrl+Esc, and Alt+Esc. It disables Ctrl+Alt+Del, but neither of the other combinations. I have a compiled executable is you want to see it. ****************README*********************************************** The pmgen.exe (32-bit) creates a simple PM window and on button 2 up tries to disable the session manager hotkeys. Cntl-Alt-Del is disabled but not alt-esc, cntl-esc etc. ******************** .C FILE *********************************** /* this one w/o migration library - like their samples */ /* Just pop a window */ #define INCL_BASE #define INCL_WIN #define INCL_DOSMEMMGR #define INCL_WINSWITCHLIST #define INCL_DOSPROCESS #include #include MRESULT ClientWndProc (HWND, USHORT, MPARAM, MPARAM) ; HAB hab ; HWND hwndClient; HSWITCH AddToSwitchList(HWND hwndFrame, CHAR *progname) { SWCNTRL swctl; HSWITCH hswitch; PID pid; WinQueryWindowProcess(hwndFrame, &pid, NULL); swctl.hwnd = hwndFrame; /* window handle */ swctl.hwndIcon = 0L; /* icon handle */ swctl.hprog = 0L; /* program handle */ swctl.idProcess = pid; /* process identifier */ swctl.idSession = 0L; /* session identifier */ swctl.uchVisibility = SWL_VISIBLE; /* visibility */ swctl.fbJump = SWL_JUMPABLE; /* jump indicator */ strcpy(swctl.szSwtitle,progname); /* program name */ hswitch = WinAddSwitchEntry(&swctl); return hswitch; } VOID PopUp(PCHAR pTxt, PCHAR pTitle) { WinMessageBox(HWND_DESKTOP, HWND_DESKTOP, pTxt, pTitle, 1,MB_ENTER); } struct _HotKeyBuf { USHORT HotKey; CHAR make_code; CHAR break_code; USHORT ID; }; int ToggleHotkey(void) { /* DosDevIOCtl performs control functions on a device specified by an opened device handle. */ HFILE DevHandle; ULONG ulCategory; ULONG ulFunction; PVOID pParmList; ULONG ulParmLengthMax; PULONG pParmLengthInOut; PVOID pDataArea; ULONG ulDataLengthMax; PULONG pDataLengthInOut; APIRET rc; /* Return code */ HFILE fh1; ULONG action; struct _HotKeyBuf HotKeyBuf[1]; rc = DosOpen((PSZ)"KBD$", (PHFILE) &fh1, (PULONG) &action, 0L, FILE_NORMAL, FILE_OPEN, /* open file */ OPEN_ACCESS_READWRITE |OPEN_SHARE_DENYNONE, /* read-only no-share 0L); if (rc) { PopUp("Failed to open KBD$ ret = %d",""); return (1); } else { PopUp("KBD opened OK\n",""); } HotKeyBuf[0].HotKey = 0; HotKeyBuf[0].make_code = 0; HotKeyBuf[0].break_code = 0; HotKeyBuf[0].ID = 0xFFFF; rc = DosDevIOCtl(fh1, 0x4, 0x56, HotKeyBuf, sizeof(struct _HotKeyBuf), NULL, NULL, 0, NULL); if (rc) { PopUp("DevIoCtl failed ret=%d",""); return(1); } else { CHAR buf[80]; PopUp("Toggled Hot Keys\n",""); } return 0; } void main (void) { static CHAR szClientClass [] = "Gen" ; static ULONG flFrameFlags = FCF_TITLEBAR | FCF_SYSMENU | FCF_SIZEBORDER | FCF_MINMAX | FCF_HIDEBUTTON | FCF_SHELLPOSITION ; HMQ hmq ; HWND hwndFrame ; QMSG qmsg ; hab = WinInitialize (0) ; hmq = WinCreateMsgQueue (hab, 0) ; WinRegisterClass (hab, szClientClass, (PFNWP)ClientWndProc, CS_SIZEREDRAW, hwndFrame = WinCreateStdWindow (HWND_DESKTOP, WS_VISIBLE, &flFrameFlags, szClientClass, 0L, 0L, 0L, 0, &hwndClient) ; AddToSwitchList( hwndFrame, "GENERIC"); // FCF_SWITCHLIST must be omitted while (WinGetMsg (hab, &qmsg, 0L, 0, 0)) WinDispatchMsg (hab, &qmsg) ; WinDestroyWindow (hwndFrame) ; WinDestroyMsgQueue (hmq) ; WinTerminate (hab) ; DosExit( EXIT_THREAD, 0); return ; } MRESULT ClientWndProc (HWND hwnd, USHORT msg, MPARAM mp1, MPARAM mp2) { switch (msg) { case WM_CREATE: return (MRESULT) 0; case WM_ERASEBACKGROUND: return (MRESULT) TRUE; case WM_BUTTON2UP: ToggleHotkey(); break; } return WinDefWindowProc (hwnd, msg, mp1, mp2) ; } # This is just like samples (with /Ti+ & /O- cp = icc -c -Gd- /Se /Re /ss /Ms /Gm+ /Ti+ /O- LIBS = DDE4MBS os2386 .c.obj: $(cp) $*.c lp = /NOD /De /CO switch.exe: pmgen.obj link386 $(lp) $**,pmgen.exe,NUL,$(LIBS),pmgen.def; echo make done *************.DEF FILE***************************** NAME PMGEN WINDOWAPI DESCRIPTION 'OS/2 Presentation Manager Graphics Demo' PROTMODE STUB 'OS2STUB.EXE' STACKSIZE 8192 The attached test case fails to woak as documented. The DevIOCtl a Category 4, function 56H and a HotKey ID of -1 is supposed to disable Ctrl+Alt+Del, Ctrl+Esc, and Alt+Esc. It disables Ctrl+Alt+Del, but neither of the other combinations. I have a compiled executable is you want to see it. Refer to Physical DD book, page 18-81, 18-82. ****************README*********************************************** The pmgen.exe (32-bit) creates a simple PM window and on button 2 up tries to disable the session manager hotkeys. Cntl-Alt-Del is disabled but not alt-esc, cntl-esc etc. ******************** .C FILE *********************************** /* this one w/o migration library - like their samples */ /* Just pop a window */ #define INCL_BASE #define INCL_WIN #define INCL_DOSMEMMGR #define INCL_WINSWITCHLIST #define INCL_DOSPROCESS #include #include MRESULT ClientWndProc (HWND, USHORT, MPARAM, MPARAM) ; HAB hab ; HWND hwndClient; HSWITCH AddToSwitchList(HWND hwndFrame, CHAR *progname) { SWCNTRL swctl; HSWITCH hswitch; PID pid; WinQueryWindowProcess(hwndFrame, &pid, NULL); swctl.hwnd = hwndFrame; /* window handle */ swctl.hwndIcon = 0L; /* icon handle */ swctl.hprog = 0L; /* program handle */ swctl.idProcess = pid; /* process identifier */ swctl.idSession = 0L; /* session identifier */ swctl.uchVisibility = SWL_VISIBLE; /* visibility */ swctl.fbJump = SWL_JUMPABLE; /* jump indicator */ strcpy(swctl.szSwtitle,progname); /* program name */ hswitch = WinAddSwitchEntry(&swctl); return hswitch; } VOID PopUp(PCHAR pTxt, PCHAR pTitle) { WinMessageBox(HWND_DESKTOP, HWND_DESKTOP, pTxt, pTitle, 1,MB_ENTER); } struct _HotKeyBuf { USHORT HotKey; CHAR make_code; CHAR break_code; USHORT ID; }; int ToggleHotkey(void) { /* DosDevIOCtl performs control functions on a device specified by an opened device handle. */ HFILE DevHandle; ULONG ulCategory; ULONG ulFunction; PVOID pParmList; ULONG ulParmLengthMax; PULONG pParmLengthInOut; PVOID pDataArea; ULONG ulDataLengthMax; PULONG pDataLengthInOut; APIRET rc; /* Return code */ HFILE fh1; ULONG action; struct _HotKeyBuf HotKeyBuf[1]; rc = DosOpen((PSZ)"KBD$", (PHFILE) &fh1, (PULONG) &action, 0L, FILE_NORMAL, FILE_OPEN, /* open file */ OPEN_ACCESS_READWRITE |OPEN_SHARE_DENYNONE, /* read-only no-share 0L); if (rc) { PopUp("Failed to open KBD$ ret = %d",""); return (1); } else { PopUp("KBD opened OK\n",""); } HotKeyBuf[0].HotKey = 0; HotKeyBuf[0].make_code = 0; HotKeyBuf[0].break_code = 0; HotKeyBuf[0].ID = 0xFFFF; rc = DosDevIOCtl(fh1, 0x4, 0x56, HotKeyBuf, sizeof(struct _HotKeyBuf), NULL, NULL, 0, NULL); if (rc) { PopUp("DevIoCtl failed ret=%d",""); return(1); } else { CHAR buf[80]; PopUp("Toggled Hot Keys\n",""); } return 0; } void main (void) { static CHAR szClientClass [] = "Gen" ; static ULONG flFrameFlags = FCF_TITLEBAR | FCF_SYSMENU | FCF_SIZEBORDER | FCF_MINMAX | FCF_HIDEBUTTON | FCF_SHELLPOSITION ; HMQ hmq ; HWND hwndFrame ; QMSG qmsg ; hab = WinInitialize (0) ; hmq = WinCreateMsgQueue (hab, 0) ; WinRegisterClass (hab, szClientClass, (PFNWP)ClientWndProc, CS_SIZEREDRAW, hwndFrame = WinCreateStdWindow (HWND_DESKTOP, WS_VISIBLE, &flFrameFlags, szClientClass, 0L, 0L, 0L, 0, &hwndClient) ; AddToSwitchList( hwndFrame, "GENERIC"); // FCF_SWITCHLIST must be omitted while (WinGetMsg (hab, &qmsg, 0L, 0, 0)) WinDispatchMsg (hab, &qmsg) ; WinDestroyWindow (hwndFrame) ; WinDestroyMsgQueue (hmq) ; WinTerminate (hab) ; DosExit( EXIT_THREAD, 0); return ; } MRESULT ClientWndProc (HWND hwnd, USHORT msg, MPARAM mp1, MPARAM mp2) { switch (msg) { case WM_CREATE: return (MRESULT) 0; case WM_ERASEBACKGROUND: return (MRESULT) TRUE; case WM_BUTTON2UP: ToggleHotkey(); break; } return WinDefWindowProc (hwnd, msg, mp1, mp2) ; } # This is just like samples (with /Ti+ & /O- cp = icc -c -Gd- /Se /Re /ss /Ms /Gm+ /Ti+ /O- LIBS = DDE4MBS os2386 .c.obj: $(cp) $*.c lp = /NOD /De /CO switch.exe: pmgen.obj link386 $(lp) $**,pmgen.exe,NUL,$(LIBS),pmgen.def; echo make done *************.DEF FILE***************************** NAME PMGEN WINDOWAPI DESCRIPTION 'OS/2 Presentation Manager Graphics Demo' PROTMODE STUB 'OS2STUB.EXE' STACKSIZE 8192 Screen Team Approved This submitter is intending to point out (I assume) that when a PM program does the Kbd IOCtl to turn off the hot-keys that it is possible to do a Ctrl-Esc or Alt-Esc and change the focus to another PM window. You can, and this is working as designed. If however you go to an OS/2 fullscreen session you will notice that Ctrl-Esc and Alt-Esc *are* disabled and do indeed prevent you from Hot-keying to another session. Again, this is working as designed. The reason behind this behavior is that while on the desktop PM interpets all hot-keys itself and the Kbd driver does nothing more essentially than pass them along... In a full screen session however the Hot-keys are reported to the session manager which then switches sessions. Please give me a call if have any further questions... Rob x3-XXXX ═══ 2.345. 63005 BETA: Install on Mod95 does not ask for disk 6, A: noduplicate ═══ name 63005 reference abstract BETA: Install on Mod95 does not ask for disk 6, A: noduplicate state verify severity 3 answer chkm symptom use duplicate defects: none. duplicate features: none. notes: Installing on Mod 95, after configuration reboot and install options are finished, select install, says ready to install, hit enter and A: is checked and no INSERT DISK 6 msg appears, get a A: NOT READY window instead. 13-W0240. Screen Team Approved Added code to move the cdrom transfer code into the main loop of install. It will not look for the files on the diskette, but rather read the file to determine where the files are. ═══ 2.346. 63015 MM_QUERYSELECTITEM returns bogus value ═══ name 63015 reference abstract MM_QUERYSELECTITEM returns bogus value duplicate state returned severity 3 answer invtcp symptom thipe duplicate defects: none. duplicate features: none. notes: I have a sample that demonstrates the following symptons: When the mouse pointer is over the drop down menu of the app window, there is no problem. But when it is over the drop down menu of another app a WinSendMsg causes the system to hang. (must power off) My own debugging revealed that a MM_QUERYSELITEMID returns a bogus value from the other app window when there is no item selected. It returns a -1 from the local app window as it should. This is a lot easier to explain when you see it running. I have it running on my system and will be glad to demo. Red Viscuso 2-XXXX Hi Charlene, this is apparently a defect against menus. I think you'll need to ask Red for a lot more info but I am routing to the controls component for your attention, cheers, Kelvin Screen Team Approved This defect is the result of incorrect application code. For the case that causes the problem: the application tries to make a WinSendMsg call to a menu in another app. It passes the address of a structure to be filled in. Since the app does not own the memory containing the structure, it cannot write to it and a trap occurs. ═══ 2.347. 63016 ACCEL_TOGGLE Command Is Not Documented In Help Text ═══ name 63016 reference abstract ACCEL_TOGGLE Command Is Not Documented In Help Text duplicate state open severity 4 answer symptom use duplicate defects: none. duplicate features: none. notes: I talked to Jason Crawford and he told me about the ACCEL_TOGGLE command that is available for users to get access to CUA accelerator keys. This command is not documented in the on-line help. Screen Team Disapproved ═══ 2.348. 63018 BETA: 1.3 regression - DosMonWrite no longer works ═══ name 63018 reference abstract BETA: 1.3 regression - DosMonWrite no longer works duplicate state returned severity 3 answer invunr symptom fnwad duplicate defects: none. duplicate features: none. notes: This is 1.3 code that worked in 2.0, fails under the 2.1 beta. Customer is using DosMonWrite to write a character to the monitor. The call looks like: rc = DosMonWrite((PBYTE)pmonout, (PBYTE)&stRemoteKey, sizeof( stRemoteKey )); The databuffer is a structure that looks like: struct s_KeyPacket { USHORT fusMonFlags; KBDKEYINFO kbci; USHORT fusKbdDDFlag; } The call is made twice. For the first call to DosMonWrite, the contents of stRemoteKey are fusMonFlags - 0x0080 kbci.chChar - 0x64 kbci.chScan - 0x20 kbci.fbStatus - 0x40 kbci.bNlsShift - 0x00 kbci.fsState - 0x0080 kbci.time - 0x00520561 fusKbdDDFlag - 0x0000 For the 2nd call to DosMonWrite, the contents of stRemoteKey are the same except for: fusKbdDDFlag - 0x0040 Under 1.3 and 2.0, this code caused a lower case 'd' to be printed to the screen. (The 1st call is a key-down, the 2nd a key-up, since some apps need both to read the character). The goal is to insert the same messages into the system that would be sent in response to a keypress, in this case of a lower-case 'd'. Again, under 2.1 nothing appears, even though this worked under 2.0 and 1.3. Dan Lee 2-XXXX Gordon or Robert, please look at this one. Screen Team Approved All of our test cases and utility programs run just fine with the latest beta build. A utility program (M.EXE) that converts mouse movements into cursor keys and PgUp/PgDns works too, and it makes heavy use of all the DosMon functions. If user still has problem, see if he can get us a test case that reproduces the problem. --Rob x3-0730 ═══ 2.349. 63048 BETA: In 1024x768x16 ALL folders appear white ═══ name 63048 reference abstract BETA: In 1024x768x16 ALL folders appear white duplicate state canceled severity 3 answer invwad symptom vis duplicate defects: none. duplicate features: none. notes: If XGA is installed in 1024x768x16 mode, the folders and templates are not there default colors. If you change resolutions to 640x480x256 they are correct. Happens on Both IBM and Radius XGA-2 cards. Available machine in 13-W0240 rerouting per Joe Celi. -Sheila 2-XXXX Screen Team Approved The famous white icons problem, After further investigation, it has been found that the white icon problem is not a bug. When the XGA adapter has only 1/2 Mb of VRAM, 256 colors cannot be achieved at a resolution of 1024x768. Under this condition 16 colors are used. On the above system, the 8514 icon format is the one chosen by the operating system. The 8514 icon format expects 256 colors and does not use dithering. The yellow used is a very pale yellow which is not available in the physical device palette when using 16 colors. The pale yellow in the icon will be nearest fit to white when the icon is drawn to the screen. This problem will not be seen on VGA systems running with 16 colors because the icon format chosen will be the pre-dithered VGA 16 color format. The VGA format cannot be used on the XGA system running in high res because of the physical dimension differences. The VGA icon format is 32x32 and the 8514 icon format is 40x40. One possible, (but not practical) fix to this problem is to enhance the icon editor to support another device format. This icon format would use a 40x40 grid and only 16 colors. This would effect every icon in the system. All icons would grow in size because of this. The system object has been enhanced to allow the user to switch resolutions. If the user is color conscience, they can always switch to 640x480x256 colors using the system object screen resolution page. This mode is supported with 1/2MB of VRAM. Regards, Joe Celi (3/2/92) This defect is canceled due to FWAD. I think that the approximation that the 16-color driver makes in Hi-Res should not be WHITE, but Yellow or something closer to the Folder color. ═══ 2.350. 63122 BETA: DOS_BACKGROUND_EXECUTION off, video jumbled DOS sessiduplicate ═══ name 63122 reference abstract BETA: DOS_BACKGROUND_EXECUTION off, video jumbled DOS sessiduplicate state closed severity 3 answer chkm symptom vis duplicate defects: none. duplicate features: none. notes: If a DOS session is started with background_execution set to off the video loses resolution after task switch, until refreshed. In a WINOS2 session the entire pallette changes including the mouse pointer becoming a box. Any ALT-ESC switch one at time leaves the desktop unchanged. Multiple ALT_ESC to change from PM-to-DOS and back causes PM desktop to lose pallette control as well. Mod 95 setup with testcase available on debug terminal in 13-W0240. Screen Team Approved As per conversation with Susan Yee, this problem belongs to VVID Need to continuously check to see if the VDM is still frozen and issue the appropriate amount of thaws ═══ 2.351. 63241 BETA:DOS_Version setting wrong in REXX created program objeduplicate ═══ name 63241 reference abstract BETA:DOS_Version setting wrong in REXX created program objeduplicate state canceled severity 3 answer invtcp symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: Internal user submitted following problem and I reproduced it on 6.498a and GASP. It works correctly on GA. The DOS_Version value in the DOS settings in GA is 123DOS.EXE,3,40,255. In GASP and later, it is one line for each value- 123DOS.EXE 3 40 255 Bob Rosenman 3-XXXX /* */ Call RxFuncAdd 'SysLoadFuncs', 'RexxUtil', 'SysLoadFuncs' Call SysLoadFuncs classname='WPProgram' title='123_V31' location='' dos123="123DOS.EXE,3,40,255" setup='OBJECTID=;EXENAME=w:\123v31\LAN123OS.bat;PROGTYPE=VDM;MAXIMIZED=YES;SET DOS_VERSION=123DOS.EXE,3,40,255' call BldObj exit BldObj: rc= sysCreateObject(classname, title, location, setup, 'R') IF rc=1 then SAY "Programmausf}hrung erfolgreich|" ELSE SAY "Programm konnte nicht ausgef}hrt werden|" return Screen Team Approved The reason this worked in GA is because we had a bug where commas were treated as commas always and there was no way to set multiple dos versions or multiple dos devices (shown with line feeds in the dos settings dialog). The way this was fixed is as follows: when you specify a dos setting for a setup string, use comma to represent a separate setting for the keyname (ex. SET DOS_DEVICE=DEVICE1,DEVICE2) and a comma with a '^' in front to represent just a plain comma. (ex. SET DOS_VERSION=123DOS.exe^,3^,40^,255) So the test case in this defect should have '^'s in front of every comma that was intended to be an actual ',' in the dos setting. --Mindy Pollack ═══ 2.352. 63290 BETA: parameter settings disappear on reboot ═══ name 63290 reference abstract BETA: parameter settings disappear on reboot duplicate state canceled severity 3 answer valunr symptom vis duplicate defects: none. duplicate features: none. notes: This was tried on settings of \os2\cmd.exe. DO: OS/2 system; Drives icon; drive where OS/2 resides; open OS/2 folder; click on CMD.EXE with RMB to open settings; enter % in parameter field; close all of these windows. Go to Desktop Settings Notebook and add a Prompt menu to the Desktop Menu. Use Command Exe as the title, and \os2\CMD.EXE as the program to execute. If the % sign is present in the parameter field when this new menu selection then NO error msg is displayed when the window is started. If the % sign is not present an error msg is first displayed and then the window starts. However - on a reboot the parameter field was cleared and the error message was once again present when the prompt selection was started. This scenario did not occur if I used the command prompt icons to add prompt selections to the menu. The % sign in every case ( DOS full, DOS window, OS/2 full, OS/2 window ) remained after a reboot. Screen Team Approved I tried this on 6.498b and it seems to work fine. Reboot. use prompt on menu - window popsup with no error. Check the parameters of cmd.exe's file object and it still contains % I'll probably return unreproducible, but i'd first like to see it fail. I'll call Richard. Couldn't reproduce on 6.498b. Jon Grimm 2-XXXX I was able to reproduce the error on 6F.498. Notified Jon Grimm that I would show him the error on my terminal. Richard and I tried to reproduce this on 6.498G in my office and in their lab. Also tried 6.498F in their lab. Still unable to reproduce this to debug. Jon Grimm 2-XXXX ═══ 2.353. 63343 Large Rex Programs Do Not Erase Unused Extended Attributes ═══ name 63343 reference abstract Large Rex Programs Do Not Erase Unused Extended Attributes duplicate state open severity 4 answer symptom use duplicate defects: none. duplicate features: none. notes: Note: I have talked with Jeff Gray at IBM Endicott about this Rexx Defect. Mike Jasinski ----- OS2BETA FORUM appended at 15:57:48 on 93/03/03 GMT (by V$IBITG7 at BCRVMPC1) Subject: Rexx and EA's Ref: Append at 01:11:29 on 93/03/02 GMT (by KOLSKY at YKTVMH) Amir, The limit of 64KB for a tokenized image to get stored in an extended attribute is documented in the OS/2 2.0 Procedures Language 2/Rexx Reference pgs 2-18, 2-19 "Running a Rexx program". The problem is really a design issue, not a Rexx problem. The behavior in question is why doesn't Rexx clear out the unused extended attributes of a Rexx program whose tokenized image is greater than 64KB. When Rexx runs a rexx cmd file, it stores a tokenized image in the file's EAs (on an HPFS system), unless the program is so large that the image won't fit in the EA space. When this happens, Rexx can no longer store and re-use the image in the file's EA, so it must re-tokenize each time the program is re-run. The issue then becomes why Rexx doesn't release the EA space is doesn't use and I believe this is a design issue, not a defect. This is primarily due to the documented restriction in the Rexx reference and the recommendation for better performance to break up large Rexx programs such as these. The test case that was given has an error, the 7th line should be 'call a >> c3000.bfr' , not 'c3000.pre'. The test case is showing on an HPFS system that a tokenized image is stored in a file's EA, and when a rexx program is made large enough(ie. c3000.bfr), the image still exists even though Rexx no longer uses it. This situation works the same way on GA, servicePak, and the latest BORG driver. Mike Jasinski Boca Technical Interface Group ----- OS2BETA FORUM appended at 19:34:20 on 93/03/03 GMT (by HALLERMA at BCRVMPC1) Subject: Rexx and EA's Ref: Append at 15:57:48 on 93/03/03 GMT (by V$IBITG7 at BCRVMPC1) No, you should log this as a defect. Failure to free up unused or discarded space is a performance/dasd space defect just like the ram memory leak defects. Mel Hallerman Screen Team Disapproved ═══ 2.354. 63426 BETA: VXGA.SYS not found after Selective Instl. on RADIUS duplicate ═══ name 63426 reference abstract BETA: VXGA.SYS not found after Selective Instl. on RADIUS duplicate state returned severity 3 answer valunr symptom use duplicate defects: none. duplicate features: none. notes: The following problem is happening only on non-IBM systems: Systems with a RADIUS XGA-2 card work fine using XGA resolution. If a selective install is done to change video mode and XGA is selected, then when rebooting and error occurs loading VXGA.SYS in CONFIG.SYS. The file VXGA.SYS does not exist in the harddrive anymore. The file is copied back in the harddrive after doing a selective install for the DOS component. Screen Team Approved Not reproducable in 6.498. ═══ 2.355. 63466 BETA: Font reverts back to sys proportional after reboot ═══ name 63466 reference abstract BETA: Font reverts back to sys proportional after reboot duplicate state working severity 3 answer chki symptom vis duplicate defects: none. duplicate features: none. notes: Open the font palate and OS/2 System Window on desktop. Select Times Roman 8pt with LMB. Using Alt-RMB drag Tms Rmn font onto the System Menu Button attached to the OS/2 System Frame Window. This will globally change the font for pulldown and popup menus to Times Roman which it infact does. The popup menu for the OS/2 desktop is changed also as expected. But if we shutdown and reboot the font for the desktop popup menu has reverted back to the default System Proportional Font. Screen Team Approved Changed the wasy we set the presentation parameter in the LoadContextMenu procedure. We now query the profile string using the global font variables and then set the presentation parameters appropriately. Previously we "mucked" around until we got the reight pres params. (?) Roberta 3-XXXX ═══ 2.356. 63470 BETA:Project Scheduler 5 Hangs ═══ name 63470 reference abstract BETA:Project Scheduler 5 Hangs duplicate state canceled severity 3 answer invextsw symptom thipe duplicate defects: none. duplicate features: none. notes: BETA:User has reported this error. Gordon Webster has been in contact with this user and is aware of this problem. Subject: Project Scheduler 5 Hangs with OS/2 2.1 BETA From: Peter Jakab ======================================================================== Project Scheduler 5 from SCITOR corporation hangs when exiting the program. ----- OS2BETA FORUM appended at 18:07:43 on 93/01/18 GMT (by JAKAB at TOROLAB2) Subject: Hang while exiting full screen OS/2 application I'm running OS/2 2.1 BETA. Application = Project Scheduler 5 (PS5XL) / SCITOR Corporation (415)570-XXXX This application runs on a full screen OS/2 session. It uses it's own graphical interface, printer drivers and provides a Project Management Application. Everything runs fine, I can perform all functions within the program, even switch to the PM desktop and other applications. When I select the Program Exit selection from their pull down, the session hangs and I CANNOT return to the desktop, their screen is still displayed. If I hit Ctrl-Esc and Alt-Esc enough times the PM desktop will eventually come back. The ICON from which the application was started still shows active. If I try to close the application from the Task List, the system hangs. The same happens if I try to Shutdown. It's BRS time. The same scenario works just fine with OS/2 2.0 GA. PS: The program is LAN installed and runs from the LAN. I'm willing to reproduce, trace, etc. Please contact me for any further action on my part. Thanks... Peter M. Jakab 3R/312/1150/TOR JAKAB at TOROLAB2 Screen Team Approved The problem was caused by a change made in our code that was related to a defect fix in the DosClose code The Scitor mouse support DLL is multi-threaded and is using the API MouReadEventQue with the wait option. This causes that thread to block waiting for a mouse event at the Mouse device driver. A seperate thread uses MouClose when the application is exiting. This is where the hang occurs because the changes to DosClose are such that the new logic checks to see if the associated handle is in use by any other thread, and if it is in use, the DosClose thread loops waiting for the handle to be released. This caused an indefinite loop because the handle was in use by the MouReadEventQue thread which was blocked at the Mouse DD waiting for an event. The deadlock will occur in OS/2 1.3 and 2.1 but not in 2.0 or the service pack. Scitor is adding version checking code to adapt to the different problems in the different releases. In both 1.3 and 2.1, the application should not use MouClose and let the system clean up the in use resources. In 2.0, a MouClose is still necessary because the next application who tries to use mouse support in the same screen group where Scitor Project Manager was run, will not have mouse support. Gordon Webster 443-XXXX ═══ 2.357. 63620 BETA:Crosshatched Seamless Window Apps not really running ═══ name 63620 reference abstract BETA:Crosshatched Seamless Window Apps not really running duplicate state returned severity 3 answer invtcp symptom fnwad duplicate defects: none. duplicate features: none. notes: Created by CMVC to address defect 62095. DOS Compatibility MODE opens are not allowed now at startup time. CMVC:Fixed incorrect AFLD2 field. BETA: Numerous beta users have reported incidents where they have cross-hatched Window applications but they don't see them running nor are they in the Window List. I can now reproduce this problem consistantly. Drag a program template to the startup folder and give it a path to the clock.exe window application. On session page, indicate separate session. Shutdown and boot. The clock is started ok. Now drag a program icon to desktop and give it the path to notepad.exe window application. On session page , indicate separate session. Start the notepad. Shutdown with the clock and notepad active and then boot. You should get a System Error Program popup with many black holes where information should be. Sometimes this message is about the notepad app and sometimes about the clock. The message says- The process cannot access the file because it is being used by another process. If I click on okay button. I will have cross hatching on both apps, but only one is started and in the Window List. Looks like a timing problem somewhere. Running on mod 95 with VGA attached. I am not sure this is a WPSH problem. Please transfer it to appropriate component if I have picked the wrong one. Workplace is just starting the applications in a sequence that causes a file access clash. There shouldn't be a difference what order or how fast we start windows apps in separate sessions. I think you guys should look at this one first. I don't see how this is a workplace problem. --Mindy Screen Team Approved These defects are being reassigned to Ivan Lomelli. As per his request, this defect was reassigned to Carlos Romero. This problem is fixed, it needs to be verified in 6.500 rather. ═══ 2.358. 63644 keyboard doesn't function with seamless Winos2 session minimize ═══ name 63644 reference abstract keyboard doesn't function with seamless Winos2 session minimize duplicate defect 63891 state returned severity 3 answer duplicate symptom use duplicate defects: none. duplicate features: none. notes: If a seamless windows application, I tried a few, Excel 4.0, Quicken 2.0 .... is minimized, the keyboard will no longer function for navigation around the desktop. As soon as one uses the mouse to click on anything, keyboard functions are returned. Screen Team Approved I am rerouting this defect to viow after testing a private from Glenn Puchtel for defect 63891. Since this is the same problem as 63891 and was written before, 63891 is a dupe of 63644, thus the reroute. If you have any questions, please contact me, Aaron Reed, at 2-XXXX The fix for defect 63891 also fixes defect 63644, please verify on driver 6.506 which is currently being built. ═══ 2.359. 63668 BETA: PMSEEK has scrolling problems ═══ name 63668 reference abstract BETA: PMSEEK has scrolling problems duplicate state open severity 4 answer symptom fnwad duplicate defects: none. duplicate features: none. notes: This problem is believed to be a functional problem that gets manifested in PMSEEK. It might not be an error in PMSEEK but an error in the part of PM that decides when the horizontal slider bar becomes "live." Problem is in the window containing "hits." If the result is "just right" in that the last character of the filename lies on the edge of the window, you cannot scroll the window horizontally to see all of that last character. For example, I have several files named "E:\OS2BBS\OS2BETA.931" (and 932, 933, 934, etc.). If I set the search argument for "Drive E" and "*.93*" then the last digit is unreadable. And like I said, the horizontal slider is inoperative. A BETA TESTER If you need any assistance in recreating this problem, please contact Sheila McDonald-Rich on 3-XXXX. Thanks, Screen Team Disapproved ═══ 2.360. 63672 BETA:OS/2 exception popups are not properly disabled ═══ name 63672 reference abstract BETA:OS/2 exception popups are not properly disabled duplicate state open severity 3 answer symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: Problem reported by internal user Mike Miller and reproduced here by myself and Glenn Brew in the current driver at this time. Glenn asked that a defect be written. Glenn has the testcase that shows the problem. The problem is that an application written for OS/2 1.3 properly disables the SYS3175 popup when an exception occurs, but the exception popup is not disabled when the same application is run under OS/2 2.x. Screen Team Approved ═══ 2.361. 63723 BETA:Part of rectangle leftover on screen ═══ name 63723 reference abstract BETA:Part of rectangle leftover on screen duplicate state canceled severity 4 answer intit symptom vis duplicate defects: none. duplicate features: none. notes: BETA:User reported this on 6.479c and I reproduced on 6.498D. I think this is a minor problem and should be TRS'ed. Some garbage is left on the screen as result of the following scenario. A refresh fixes it. ----- OS2BETA FORUM appended at 08:43:27 on 93/03/08 GMT (by BOFRIIS at DKIBMVM2) Subject: Workplace shell/pm update prob. Ref: none. 1) press mb1 and mb2 to popup window list window 2) move the pointer outside window list window. 3) press mb1, and drag a rectangle from outside the window list window and into/over window list window, This action needs to be done faster than the system removes window list window (on my 16MHz you have about ╝ second) 4) release mb1 Now you see the rest of the rectangle (behind the window) not being deleted The leftover rect can be deleted be choosing refresh from the desktop popup, but it should NOT be necessary It is this kind of visible error which irritates the user. It makes mess on the GUI and seems so simple to fix. It makes one think, if OS/2 has this kind of problem it must be full of errors| Use Mcrsft strategy... CORRECT VISIBLE ERROR FIRST. OS/2 needs to 100% bug-free to win. Regards Jens Bo Friis, BOFRIIS at DKIBMVM2 Screen Team Disapproved This is not a borg ship issue. We will consider for the next release. -Sheila 2-XXXX ═══ 2.362. 63729 BETA:CHKDSK white output text not working ═══ name 63729 reference abstract BETA:CHKDSK white output text not working duplicate state open severity 4 answer symptom vis duplicate defects: none. duplicate features: none. notes: BETA:User submitted this and I reproduced the problem on 6.498d. All text is set white and all background is set black through the scheme palette. CHKDSK is run from the Drive A object. The variable data for type of file,total disk space and space available does not show. It should be coming out in white on a black background. (I also see some extra things that don't appear with the default color scheme.) ----- OS2BETA FORUM appended at 17:12:07 on 93/03/08 GMT (by FOREMAN at GDLVM7) Subject: CUA conformance - OS/2 breaks its own rules! Ref: Append at 15:06:15 on 93/03/08 GMT (by MJDAVID at BCRVMPC1) Looks like CHKDSK has the same problem. Here's how to see it: Using the Scheme Palette, 1) set all background colors to BLACK, 2) set all TEXT colors to WHITE, 3) ALT-drag the scheme to the desktop 4) RMB on DRIVE A, 5) select CHKDSK. When the panel with %used, etc appears, it will have the ORIGINAL BLACK TEXT, on the NEW BLACK BG, which is thus invisible. -------- For ref, the original PMR was 4X450, B916. D. J. (I want to have it MY way!) Foreman Screen Team Disapproved ═══ 2.363. 63730 BETA:Wrong help info for EDIT COLOR in EDIT SCHEME ═══ name 63730 reference abstract BETA:Wrong help info for EDIT COLOR in EDIT SCHEME duplicate state working severity 3 answer doci symptom use duplicate defects: none. duplicate features: none. notes: BETA:In trying to reproduce a user problem, I ran into this problem with help information. Open Scheme Palette,click on edit scheme,click on help,double click on edit color. This help panel has steps 1-4 which can not be done when you get to this help panel from the scheme palette. This panel is also used when you get to it from the color palette and in that case, the info is correct. It is not correct when you get to it from the scheme palette and the steps can not be done. (tells how to drag colors with mouse) Screen Team Approved ═══ 2.364. 63765 SETBOOT - Undocumented Switches ═══ name 63765 reference abstract SETBOOT - Undocumented Switches duplicate state working severity 3 answer chki symptom use duplicate defects: none. duplicate features: none. notes: The following switches for SETBOOT are not documented in the online reference. /IBA, /IBOOTA, Working. Vince Spadea. Target 3/22/93. Screen Team Approved ═══ 2.365. 63772 BETA: DOS sessions only ═══ name 63772 reference abstract BETA: DOS sessions only duplicate state working severity 3 answer doci symptom fnwad duplicate defects: none. duplicate features: none. notes: ===== OS2BETA CFORUM created at 20:14:53 on 92/12/28 GMT (by HOPKINS at AUSVM8) In the on-line Command Reference, expand "Information about OS/2 Commands," choose "How to Read a Systax Diagram," then select "DOS sessions only." There it states "When using the OS/2 operating system, you may choose to use the DOS environment similar to IBM Disk Operating System Version 4.0." ~~~ Later, it says "You can start DOS tasks or begin a DOS program almost as though you were using DOS 4.0." ~~~ Say what? That explains a whole lot of appends about how DOS support differs from DOS 5.0! (Who in his right mind would want to emulate DOS 4.0?) A Beta Tester If you have any questions, please contact Sheila McDOnald-Rich on 3-XXXX. Thanks... Screen Team Approved ═══ 2.366. 63774 BETA: OS/2 sessions only ═══ name 63774 reference abstract BETA: OS/2 sessions only duplicate state working severity 3 answer doci symptom fnwad duplicate defects: none. duplicate features: none. notes: In the on-line Command Reference, expand "Information about OS/2 Commands," choose "How to Read a Systax Diagram," then select "OS/2 sessions only." There it states "... This gives each program the potential of taking advantage of available physical memory up to 16MB." ~~~~ Not too impressive ;-) ~~~ A Beta Tester If you have any questions, please contact Sheila McDOnald-Rich on 3-XXXX. Thanks... Screen Team Approved ═══ 2.367. 63775 BETA: WIN-OS/2 File Manager -- Disk Options ═══ name 63775 reference abstract BETA: WIN-OS/2 File Manager -- Disk Options duplicate state open severity 4 answer symptom fnwad duplicate defects: none. duplicate features: none. notes: I should add... there are a lot of words in the Help for File Manager about how to do things like format a diskette. Better clean that stuff out, I think (whatever you don't do/support, then don't talk about it). Since these options are grey out in the WIN-OS/2 File Manager's Disk pull-down, they need to be removed from the help text as well. A Beta Tester If you have any questions, please contact Sheila McDOnald-Rich on 3-XXXX. Thanks... This defect is TRS'able! If you have any questions, please contact Sheila McDOnald-Rich on 3-XXXX. Thanks... This defect is being reassigned to component WINOS2. For OS/2 2.1, WINOS2 has the responsibility for updating the Microsoft helps (from Microsoft's Windows 3.1 code). Those helps are then included in the OS/2 product along with the OS/2 (Win-OS/2 emulation) helps that are written by Boca Information Development. Screen Team Disapproved ═══ 2.368. 63790 BETA:Working directory entry not honored for Windows App ═══ name 63790 reference abstract BETA:Working directory entry not honored for Windows App duplicate state canceled severity 3 answer intit symptom use duplicate defects: none. duplicate features: none. notes: BETA: A user ran into this problem and in researchimg what was happening, I spoke with Ayo Anise of winos2 and Mindy Pollack from the WPS. After speaking with them, I think there is a problem somewhere that would affect the product's usability. I wrote the following to Kerry Ortega in the Usability group and have appended his response. The problem can be reproduced by using the WRITE app that comes in winos2. I am starting this out in the WPS, but winos2, session manager and id may need to be involved. ---------------------------------------------------------------------- I think there is a usability problem with 2.1 and want to make you aware of it. Program objects have a settings notebook and one of the fields on the program page is called Working Directory. That field is used for all types of apps except Windows apps. Because of the way winos2 works, that field is ignored. For all other apps, that field is used. To an end user, the program page looks the same regardless of the type of application, but the Working Directory field is ignored if it is a Windows app. Would you support a defect to get the Working Directory field greyed if the path points to a recognized Windows app or should the field be honored? Windows 3.1 has nothing comparable to this. The working directory is set through the properties dialogue and it also works that way for winos2. ------------------------------------------------------------------------ The following response was from usability (Kerry Ortega) Ideally, the field should be honored. It that can't be done in Borg maybe this should be revisited in Star/Picard (i.e., let's not lose this problem). In the near term, greying this field would be acceptable. Greying would happen if one of the Window radio buttons were selected on the session page (either by the user or automatically by the system based on the app type). If text was already entered in the Working Directory and someone changed the type of program on the session page, I'd leave the text there and grey it. Even if we don't make a change we should document the restriction on the help panel for this page and the section in the using book about running Windows (I think it's CH12). Screen Team Disapproved This is not a borg ship issue, we will consider for the next release. -Sheila 2-XXXX ═══ 2.369. 63802 BETA: wrong or missing help info for formating optical drivduplicate ═══ name 63802 reference abstract BETA: wrong or missing help info for formating optical drivduplicate state working severity 3 answer docm symptom vis duplicate defects: none. duplicate features: none. notes: If you select the object for the optical drive and select format a dialog box appears. From this dialog box click on HELP. From the help screen double click on Long Format. The help associated with long format presently relates to FONTS and needs to be replaced with the appropriately worded help info. In addition the master help index contains no reference to locking and unlocking optical drives. For further information or for the exact wording of the help text contact Amol Khedkar at 2-XXXX he is the developer of the format command. Screen Team Approved ═══ 2.370. 63807 BETA:Cannot get to Desktop when printing in View utility ═══ name 63807 reference abstract BETA:Cannot get to Desktop when printing in View utility duplicate state open severity 3 answer symptom use duplicate defects: none. duplicate features: none. notes: BETA: User reported this problem on 6.479c and I have reproduced on 6.498d. Bring up the view utility with cmdref. (at prompt- view cmdref). Select print, then all sections, then print and yes. While it is printing, you cannot minimize the window from either that window or the window list and you cannot get to the desktop unless you cancel the print or wait till it finishes. Screen Team Approved ═══ 2.371. 63809 BETA:View print radio buttons should be greyed when printinduplicate ═══ name 63809 reference abstract BETA:View print radio buttons should be greyed when printinduplicate state canceled severity 4 answer dsamt symptom use duplicate defects: none. duplicate features: none. notes: BETA: User reported this on 6.479c and I reproduced on 6.498d. I think this problem should be TRS'ed to next release. If you bring up the view utility and request to print, the 3 print choices should be greyed when printing is occurring. They are selectable and shouldn't be. Screen Team Disapproved This will have to be done next time around. ═══ 2.372. 63816 BETA: PMChart causes SYS3175 error ═══ name 63816 reference abstract BETA: PMChart causes SYS3175 error duplicate state open severity 3 answer symptom vis duplicate defects: none. duplicate features: none. notes: Start PMChart and select the Pencil tool -1st rectangle tool, and draw a rectangle in the top half of the screen and another rectangle in the bottom half of my screen. Then select pencil tool straight line tool and draw a vertical line on the screen. Double click on system menu button to close application. When prompted to save changes click NO. Application will terminate with SYS3175 error message. Screen Team Approved ═══ 2.373. 63827 BETA: GpiPolygons online help boo-boo ═══ name 63827 reference abstract BETA: GpiPolygons online help boo-boo duplicate state working severity 3 answer doci symptom use duplicate defects: none. duplicate features: none. notes: This ought to be a quick fix: in the online help for the API GpiPolygons, the text says you have to define INCL_GPIPRIMITIVES to get the api. That won't do it though. You have to define INCL_GPIPOLYGON instead. Dan Lee Screen Team Approved The IPMH is a component for an old online help tool. It has nothing to do with the information you refer to in this defect. I am returning it so that you can find out who owns the information. You can then assign the defect to a correct component. Mike Bresnahan 2-XXXX ═══ 2.374. 63829 BETA: Wordperf - ptr handles valid after shell exit ═══ name 63829 reference abstract BETA: Wordperf - ptr handles valid after shell exit duplicate state open severity 3 answer symptom use duplicate defects: none. duplicate features: none. notes: WPS uses WinSetPointerOwner on icons that it loads to allow those pointers to be displayed from different processes, and to prevent them from being accidently destroyed. If WPS takes an exception or exits, there doesn't seem to be any code that goes back and tries to clean up those shared pointer handles. WordPerfect is using a program that starts and stops the WPS now and then, and some pointer handles are still valid AFTER the shell is no longer running. In other words, every time the WPS terminates it is leaving behind a variable number of icons that never get cleaned up by OS2. What is needed is an exit list routine that cleans up these shared resources when the shell goes down. Dan Lee Screen Team Approved ═══ 2.375. 63830 BETA: icons for programs on network are lost when rebootingduplicate ═══ name 63830 reference abstract BETA: icons for programs on network are lost when rebootingduplicate state open severity 3 answer symptom vis duplicate defects: none. duplicate features: none. notes: Icons for programs that reside on a network drive are lost when the machine goes through a normal shutdown and is then rebooted. This defect was TRS'd in 1992 against R202. Major customers including Nations Bank want this issue resolved. Please reference defects #14356 and #14899 for additional information and problem history. Screen Team Approved I tried to assign the apar PJ08131 to this defect but somehow the afld1 got updated. Please correct this thanks. ═══ 2.376. 63834 Beta: General Notebook Page -- Default ═══ name 63834 reference abstract Beta: General Notebook Page -- Default duplicate state returned severity 3 answer invwad symptom use duplicate defects: none. duplicate features: none. notes: Curiosity compelled me to try the "default" pushbutton on some of the General pages of various Notebooks. Some things changed from the way the system is distributed which prompted a "mini audit" of many more Notebooks. Examples where "nothing" changes include: Startup, Drives, Templates and Desktop. Examples where the title changes to "Folder" and the icon changes to a folder include: System Setup, Productivity, Games, Command Prompts, Information and OS/2 System. In these cases, "Undo" gets the old title and icon back. Examples of objects within the System Setup folder where nothing changes include: System, System Clock, Sound, Mouse and Country. Examples of objects within the System Setup folder where the title changes slightly ("Palette" becomes "palette") are: Scheme Palette, Color Palette and Font Palette. "Undo" restores the correct titles. Examples of objects within the System Setup folder where the title changes to "Program" are: Device Driver Install and Selective Install. "Undo" restores the correct titles. *IF* the system expects that the original titles and icons should be refreshable, should the user mess them up, then perhaps a more complete audit should be performed and the defaults set up correctly. A Beta Tester If you have any questions, please contact Sheila McDOnald-Rich on 3-XXXX. Thanks... Screen Team Approved It works as designed. We can change it in the next release. ═══ 2.377. 63836 WLOized version of Win app GrAF access violation ═══ name 63836 reference abstract WLOized version of Win app GrAF access violation duplicate state open severity 3 answer symptom thipe duplicate defects: none. duplicate features: none. notes: Running a WLOized version of GrAF, an as of yet unshipped, app by Texas Instruments causes access violation at 13fe7d81 IBMXGA32.DLL 0003:00017d81 p1=00000001 p2=13290088 p3=x p4=x EAX=13290000 EBX=00002620 ECX=00024dc8 EDX=00000050 ESI=00002620 EDI=00005c1e DS=0053 DSACC=d0f3 DSLIM=1bffffff ES=0053 ESACC=d0f3 ESLIM=1bffffff FS=150b FSACC=00f3 FSLIM=00000030 GS=0000 GSACC=* GSLIM=* CS:EIP=005b:13fe7d81 CSACC=d0df CSLIM=1bffffff SS:ESP=0053:00024d88 SSACC=dof3 SSLIM=1bffffff EBP=00024e3c FLG=00012246 I can be contacted at 2-XXXX and have the app as well as a test case set up with a debug terminal. Screen Team Approved ═══ 2.378. 63843 Beta: Preemptive Multitasking? ═══ name 63843 reference abstract Beta: Preemptive Multitasking? duplicate state open severity 4 answer symptom use duplicate defects: none. duplicate features: none. notes: I've a file that contains, give or take a little, all of the "appended by" lines and all of the "Subject: " lines of this OS2BETA CFORUM. That's about 520K of stuff (over 8000 lines). I invoke EPM against this file from the Command Line then, via the "Command dialog," search for all occurences of "Subject: " with this command: all /Subject: / Once the search is begun, the notion of "preemptive multitasking" takes a back seat to the search. Task switching? Takes forever. Start an OS/2 Window? Takes forever. My expectation was that the operating system would respond to my inputs and put the EPM search in the background. I have a little shareware program, TACH.EXE, that shows CPU percentage and the meter is pegged at 100%. Eventually, if I do not panic, EPM creates an ".ALL" file from this search of about 134K (about 3000 lines). The system is an 8MB 386/33. The SWAPPER.DAT file is 4MB and does not grow. The file I edit is on a FAT partition while the operating system is on an HPFS partition. A Beta Tester If you have any questions, please contact Sheila McDOnald-Rich on 3-XXXX. Thanks... Sheila, there is no information at all in this defect to imply that PMWIN is causing this. There is no testcase either. You need to take Beta feedback like this to the support folks who are supporting the Beta and have them review the problem. Please call me first next time. I want to see all Beta problems get fixed but this one needs investigation by the support folks first. Kelvin 3-XXXX Reading this again it sounds as though maybe you are the one experiencing the problem. You need to get the performance folks to measure the system and see what is slowing down. Kelvin. Kelvin, I am supporting the beta and have verified this problem on my personal system. It is as the user states. I can recreate this problem for you, please call me. Whether the problem exists in PMWIN is questionable. I didn't originally open this defect against PMWIN, it was assigned. Thanks This appears to be an EPM problem from the description. Kelvin 3-XXXX Screen Team Disapproved ═══ 2.379. 63891 Can not type in OS/2 window after closing Progman ═══ name 63891 reference abstract Can not type in OS/2 window after closing Progman duplicate state working severity 3 answer asni symptom use duplicate defects: duplicate features: none. notes: To recreate: 1. bring up OS/2 window 2. type 'start /win progman' 3. when progman comes up, double click to close it. 4. Hit return at close pop-up. 5. Now focus should be returned to OS/2 window, but you can't type anything. Changing focus and then restoring it to the OS/2 window fixes the problem. This problem has been visible for several months. Mark Rogalski 2-XXXX Screen Team Approved We activating a seamless "Blackhole" the frame flags where not updated (or-ed in) with "FF_ACTIVE". This caused WM_QUERYFOCUSCHANGE messages not to be sent when this window was destroyed, thereby failing to reset focus correctly. I have a private that fixes this problem (glenn puchtel) In order to properly emulate the activation processing for a Seamless VDM, we had to tweak the frame flag so that WM_QUERYFOCUSCHAIN messages will get process correctly. ═══ 2.380. 63913 BETA: DOS window shows no update after DOS F/S task switch.duplicate ═══ name 63913 reference abstract BETA: DOS window shows no update after DOS F/S task switch.duplicate state canceled severity 3 answer valunr symptom use duplicate defects: none. duplicate features: none. notes: If you start a DOS window session and run PKZip on a bunch of files in the window, then open a DOS F/S and do the same thing. Switch back to the windowed session thru task switch and window is not updated as the PKZip continues. Click with the mouse to establish focus and it scrolls with what it was doing. not until. 13-W240 Debug available. Screen Team Approved This problem was APAR'd long time ago against VVID. The fix was never put into the build because the PTR (B736257) didn't the attention it needed at that time. VVID should screen the type of events to make certain that a repaint occurs when PM is switched back into the foreground. Problem can not be recreated with the 6.504 driver. I could not reproduce this problem on the 6.504 revision. Cancelled this Defect. ═══ 2.381. 63914 BETA: Start of DOS F/S after DOS window causes trap on #2 tduplicate ═══ name 63914 reference abstract BETA: Start of DOS F/S after DOS window causes trap on #2 tduplicate state canceled severity 3 answer invinf symptom thipe duplicate defects: none. duplicate features: none. notes: Start a DOS window session, then a DOS full Screen session. Close both from desktop,cmd line or task list. Than repeat the process to get Trap. If you start the F/S first then the Windowed it does NOT occur! LN showed different thing each time I caused the trap. 13-W240 Debug available. After talking with Steve, this appears to be a DOS emulation problem... This is apparently a problem with the latest build and VKBD has been static for I'd guess the past 2 or 3 builds. --Rob x3-0730 Screen Team Approved I could not reproduce this problem, and Steve was also unable to reproduce this problem yesterday. This problem occured because of a earlier test on the same machine with a Willowized WIN app. I tried to recreate but could not. Will hold in limbo until I can verify problem or leave cancelled. . ═══ 2.382. 63946 BETA:SYS0447 keyboard focus is not available on piping MOREduplicate ═══ name 63946 reference abstract BETA:SYS0447 keyboard focus is not available on piping MOREduplicate state working severity 3 answer dsai symptom fnwad duplicate defects: duplicate features: none. notes: BETA: User reported this problem and I have recreated the problem. This problem was more consistent in previous beta driver, but is back in this one. Previous defect number was 59591. In an OS/2 window type - type config.sys | more and press enter to continue seeing the info. About 25% of the time I get a SYS0447 The keyboard focus is not available. Nothing else is running at the time. Mod 95. Screen Team Approved Changes were made to allow the calling thread to selectively grab the synchronization semaphore ONLY when the owning thread ID that owns the semaphore is different. The wait flag default was also changed from IMMEDIATE_RETURN to INDEFINITE_WAIT to allow for applications that were written that do not check return codes in error conditions. ═══ 2.383. 64016 BETA:INT 3 hit followed by jump to itself, system hang ═══ name 64016 reference abstract BETA:INT 3 hit followed by jump to itself, system hang duplicate state open severity 3 answer symptom thipe duplicate defects: none. duplicate features: none. notes: BETA: Users have reported system hangs during config.sys processing that they thought was associated with the presence of a VDISK statement in config.sys. Another appender, Larry Bank, who used to work in Technical Support has appended that the presence of VDISK statement is causing IBMVGA32.DLL to hit an INT 3, followed by a jmp to itself. The INT 3 is being generated in function 'ring3_CriticalError' which is intended to be called only when the display driver hits something absolutely fatal such that it cannot continue. An INT 3 is generated (so that if you are in a debugger you can see it) and a JMP to self to cause the system to hang (if you are not in a debugger) because the driver is unable to recover and continue. The ONLY place in the VGA driver that this occurs is during initialization (which takes place the first time the DLL is loaded by the graphics engine) as a result of receiving a bad return code from the 'Vio32SetMode' function call. It is possible that the base video handler (BVH) is returning an error because of some interaction between the VDISK device driver and video device drivers that are in the CONFIG.SYS file. This defect should be routed to base video. David A Kerr The information provided in this defect doesn't describe what were the failing configurations: VGA, SVGA or other display driver running, and what was the VDISK statement format. I contacted the submitter of the defect requesting more details, but he was not willing to cooperate, citing that it is not his responsability and that he is too busy for it. I left a message for Larry Bank, asking for more setup information. I also asked our FVT testers to create a VDISK on one of our SVGA systems and attempt to reproduce. We have tried reproducing the problem on VGA and SVGA setups (IBM model 40 used), couldn't cause the failure, even if VDISK had invalid parameters. Larry Bank's setup was a 486DX-50MHz clone with a VGA and a monochrome secondary. In his experience, placing VDISK even at the end with 4M 512 512 parameters would cause the hang. We will try to reproduce it on a Gateway 66DX2, without a secondary (since we don't have any monochromes in the lab). Screen Team Approved We attempted to recreate the problem on our VGA nad SVGA setups which were installed in IBM model 40 machines, and no problems were encountered. However, on a ALR 486DX clone, which was successfuly running SVGA 1024x768, after the VDISK.SYS was specified as the last statement in the CONFIG.SYS, the system booted, display driver set its mode (no INT 3 encountered) but a minute after this point, the system went into a boot sequence on its own. After that, it would not boot again. If OS2 booted from a diskette, dir on boot drive returns "Invalid file specified" error. Howard or Annie in out FVT lab will reinstall the ALR and help with the debugging process. I have briefly described this problem to Miguel Ascensio and he is willing to take a look at it. ═══ 2.384. 64022 FoxPro 2.5 GPF in Toolhelp.dll at 0001:2DEF ═══ name 64022 reference abstract FoxPro 2.5 GPF in Toolhelp.dll at 0001:2DEF duplicate state canceled severity 3 answer intmt symptom thipe duplicate defects: none. duplicate features: none. notes: FoxPro 2.5 for Windows GPFaults in a variety of scenario's in toolhelp.dll at 0001:2DEF. The easiest way to reproduce is to select Help\Puzzle\Shuffle, but it occours in other ways as well. This is partial fragment of TOOLHELP.DLL where the trap occurs 00002dd8 b88416 mov ax,1684 Get device API entry point 00002ddb bb0500 mov bx,0005 Virtua timer 00002dde cd2f int 2f return ES:DI 00002de0 8cc0 mov ax,es VxD entry point 00002de2 0bc7 or ax,di 00002de4 743d jz 2e23 00002de6 897efc mov word ptr [bp-04],di 00002de9 8c46fe mov word ptr [bp-02],es 00002dec b80101 mov ax,0101 00002def ff5efc call dword ptr [bp-04] call to timer function FoxPro 2.5 for Windows runs OK under standard mode. However, the traps occurs while running under enhanced mode. It happens because the current OS/2 2.1 Beta version doesn't support full enhanced mode. Any call to query the get the entry point of VxD is not fully implemented, therefore it generates a fault. Typically, we are running in a dos environment with VxD supported, but this is OS/2 VxD and not Windows VxD. The code for supporting VxD for enhanced mode will be in next release. Joe Luu 26843 ═══ 2.385. 64023 BETA:Limit of 99 for EPM no longer needed ═══ name 64023 reference abstract BETA:Limit of 99 for EPM no longer needed duplicate state working severity 3 answer docm symptom use duplicate defects: none. duplicate features: none. notes: BETA: User submitted this on OS2BETA forum. Recommend TRS . ----- OS2BETA FORUM appended at 01:19:43 on 93/03/16 GMT (by ASHAN at CANVM2) - Subject: Small User's Guide error Found this in the User's Guide: >ENHANCED EDITOR (OS/2) >o The OS/2 Enhanced Editor supports up to 99 fonts. If additional > fonts are installed, unpredictable results might occur. This is no longer true since the first EPM update (or since SP or with EPM 5.51). It now handles more than 99 fonts properly. Albert Shan OS/2 Ambassador / Vancouver ISC Level 2.5 OS/2 Support Reassigned to Karla Stagray. 3/17/93 Anita Soto Marion Lindsey requested this defect and is assigned to closure. Target 3/22/93 Screen Team Approved ═══ 2.386. 64065 LOCKUP does not activiate while users code is running ═══ name 64065 reference abstract LOCKUP does not activiate while users code is running duplicate d state e.open severity ne.3 answer symptom fnwad duplicate defects: none. duplicate features: none. notes: BETA: User reported this on 6.479c and it is still present. He has a program written for OS/2 1.3 which allows the LOCKUP function to work in 1.3 when his program is active, but LOCKUP does not work in 2.x when the same program is active. Program is called DISKG. Kim Tran has a diskette which has the executable that shows the problem and the source code for that executable. The submitter thinks that WinSetWindowText() generates a WM_MOUSEMOVE message for some reason and that may be what is causing the LOCKUP to not execute. User reported this on 6.479c and it is still present. He has a program written for OS/2 1.3 which allows the LOCKUP function to work in 1.3 when his program is active, but LOCKUP does not work in 2.x when the same program is active. Program is called DISKG. Kim Tran has a diskette which has the executable that shows the problem and the source code for that executable. The submitter thinks that WinSetWindowText() generates a WM_MOUSEMOVE message for some reason and that may be what is causing the LOCKUP to not execute. Screen Team Approved WinSetWindowText does not generate wm_mousemove message. Even when the message gets generated by adjusting the window position, LockUp does check if the mouse was actually moved. Notice that LockUp works fine when DiskG window is not minimized. The problem was discussed with R. Carey and P. Magid. The defect is routed to LockUp for further investigation. Kim Tran