OS/2 Professional

home *** CD-ROM | disk | FTP | other *** search

/ OS/2 Professional / OS2PRO194.ISO / os2 / com / bbs / btp_eebe / btee_mod.doc < prev next >

Wrap

Text File | 1992-09-12 | 45.2 KB | 1,169 lines

BinkleyTerm-OS/2 2.5 EE (Extended Edition) BETA E 12 September '92 BinkleyTerm-DOS 2.5 EE (Extended Edition) ******************************************************************** MICS OS/2 Forum BBS: ++41-41-53 86 07 (HST, V.32b) Michael Buenter ++41-41-53 86 27 (V.32b, FAX) Untere Wiese 3 FidoNet: 2:301/601 and 2:301/602 CH 6020 Emmenbruecke SIGNet: 27:1341/103 Switzerland Internet: buenter@iis.ethz.ch VERY IMPORTANT ************** PLEASE DO NOT ASK QUESTIONS OR REPORT BUGS IN PUBLIC BINKLEY ECHOS! THIS VERSION HAS NO SUPPORT FROM THE ORIGINAL AUTHORS OF BINKLEYTERM! If you don't accept this, there may be no source release of any official and "hacked" BinkleyTerm version anymore! So, if you don't agree with those two capital sentences, please delete this copy of BT EE. Thanks. I don't support this version, neither. But I'll give my best to answer your questions which arrive per NETMAIL. As this is only a Beta Release I suggest you not to make it publicly available. It's ME who has to support all twits (please consider that)... There's a beta talk and support echo available on my and many other (german) node: BTEEBETA. Just ask if you're interested in carrying it, too. The original Version of this document was made by Holger Schurig. Many thanks to him for his great effort in adding so many nice features to BinkleyTerm 2.40! Many thanks also to the Binkley Team who created a great mailer. I LOVE IT! About This Documentation ------------------------ This documention discribes the modifications made by me to BinkleyTerm 2.50. I implemented most features which Holger Schurig made in his BT 2.40 HS Version, but there are lots of new features from my side... My version announces itself as "Binkley 2.50 EE-x" (x represents the beta level), so you can differ it from the original BinkleyTerm. It is compiled it with MSC 6.0a (DOS and OS/2 16bit) or with C Set/2 (32bit OS/2 version). Everything you don't find here you can find in BT_REF.DOC, BT_USER.DOC, these files are included in the original distribution package. A draft version of an updated BT_REF.DOC is included in this archive. BT_USER.DOC will follow soon. See announcement in BTEEBETA. The Original ------------ You can request the unmodified and complete original as BEXE_250.LZH BinkleyTerm 2.50, DOS executables BDOC_250.LZH BinkleyTerm 2.50, documentation. BEXEN250.LZH BinkleyTerm 2.50, OS/2 executables and configs, without snoop server code. BEXES250.LZH BinkleyTerm 2.50, OS/2 executables and configs, with snoop server code. BSRCP250.LZH BinkleyTerm 2.50, OS/2 source modules (updates) BSRC_250.LZH BinkleyTerm 2.50, source code. My Beta Version is requestable as BTPBETA for the OS/2 2.0 32bit Version BTBETA for the overlayed DOS Version --> (available soon!) <-- BTBBETA for the nonoverlayed Version. --> (available soon!) <-- The 16bit OS/2 version is available on request. License ------- This beta-version is provided 'as is', without warranty of any kind, either expressed or implied. I only guarantee that it occupies diskspace. In no event will I be liable to you for any damages, including lost profits, lost savings or other incidental or consequential damages arising out of the use of this program. This version is free of charge and bound to no conditions other then the ones found in LICENSE.250. But one plea: if you have any ideas for improvements, please send them to me. If it is something that benefits the whole community and if i can manage to program it, I will probably implement it. Team Work --------- Comment from Holger Schurig to his 2.40 HS version: Most additions/modifications of this Binkley-Version are mine. But i am not the only one who contributed to this version, Jan Egner and Leo Moll @ 242/2 made some additions, too. Leo Moll especially implemented all of our modifications from "Binkley 2.30 (Version HS)" into the "Binkley 2.40 (Version HS"). A real huge work. Another thankyou to Hans Theo Wolf and Alex Wyss who contributed my work with many ideas. My comments: I translated Holgers additions/modifications to V2.50 EE and added a lot of new code myself. Many thanks to my beta testers, who sent in a lot of fixes and ideas. I'd like to thank especially Alex Wyss for many good ideas! Many thanks of course to Roland Gautschi! More Comments ------------- This beta is the last version which will come out for OS/2 1.x and DOS. I don't intend to develop the 8/16bit "compatibility" code any further. There won't be a 32bit new beta for the next months, either. Things are developping fast and I think it's time for the big change now. Let's see what happens. Have fun with this Extended Edition. Modifications ------------- - = modifications from V2.40 HS * = my own, new modifications - Binkley will now clear the screen upon start up. [ScreenBlank] - The "Thanks for using BinkleyTerm 2.40" and "Another fine user-DEVELOPED software package!" has been sent to limbo. [NoThankyou] - This version of Binkley uses a different method of hanging up the phone (if no special string like DialInit, HangUp etc defined): it sets DTR to false. Then it waits until the carrier is gone and one additional fifth of a second is passed. So it is something faster than the original Binkley, which always waited 2 seconds whether the carrier was present or not. [FastModem] - Together with FastModem I changed the pause which makes BT wait for an "OK". [NoOk] - The ANNOUNCE-string is changed to "BinkleyTerm 2.50 EE-Bx", to differ my version from the original. [Version] - The Colors statement in BINKLEY.CFG now works as follows: Colors Colors n1 n2 n3 n4 n5 n6 n7 n8 n9 n10 n1 top and bottom text displays (Nodenumber, Systemname, Version and help key). n2 "Current-Settings"-window n3 "Today at a Glance"-window n4 "Pending Outbound Mail"-window n5 "Recent-Activity"-window n6 "Transfer Status"-window n7 actual called node in the "Pending Outbound Mail"-window n8 color of pop-up windows n9 header of all windows n10 border of all windows [Colors] - The time will now be displayed in a standard European format: "Mo 26.09.89, 11:10" (this is: day, month, year). [Timeofday] - Normally you can execute event-scheduled programs only through ErrorLevel errorlevel exits. But MSDOS batch execution is rather slow and your "Recent Activity" window will always be cleared. ALT F1-F9 Now you can specify negative 'errorlevels' between -1 and -9 that correspond to the ALT-F1 through ALT-F9 keys you can define with "Shell <n> <command>" in your configuration file. [FastEvent] * Expanded negative errorlevel control to all places, where you can define error levels. But there are still problems with dynamic events. So you better use the "Aftermail" features! [FastEvent] - If you define a negative Log level, Binkley will only write to LogLevel the logfile if a carrier is currently detected. This reduces the logfile dramatically, but with "LogLevel -5" you still have the complete information of your sessions available. Note that this can cause problems with logfile-analyzers. [StatusLog] - The Standard Binkleyterm scans the hold directory every 10 ALT-O minutes, but only if there is no pending outbound mail present. Now you can press ALT-O if you want to rescan the hold directory, e.g. if you have linked two or more computers, and one of them has just placed a crash bundle in the hold directory of your Binkley computer. [Outbound] - You can always init your modem manually with ALT-I. Faster and ALT-I nicer then the ALT-T ALT-U sequence used by many. ALT-I won't rescan the outbound. [ReInit] - If you call the shell Alt-J and if you have defined a busy string, Binkley will now set the modem off hook. [GeneralExec] - Some people here in Germany use a hardware watchdog to reset a ReinitTime hanging system depending on 5 minutes silence on the modem line. For these (and other with unstable modems) is the new configuration statemant "ReInitTime <minutes>". If you configure "ReInitTime 40", your modem will be initialized every 40 minutes. Note that this is the same time binkley waits until blanking the screen, if you enabled screenblank. The people net using a hardware-watchdog with a retriggerable monoflop are very happy with this! [ReInit] - New statement "ReadHoldTime <minutes>" for BINKLEY.CFG. The ReadHoldTime standard Binkley reads your Hold-Directory every 10 minutes if there is no active pending mail. This statement causes Binkley to read the hold directory every <minutes> minutes, independently from the state of "Pending Outbound Mail". Note that Binkley will also read in the hold directory upon start up or restart (after shelling, eg). This is nice for multitasked or networked environments. [ReadHold] - For national purposes: new statement "CostLog <file>". It needs CostLog definitions from parselst.cfg and works only with the German and Swiss phone billing scheme. If "CostLog" isn't defined, then everything works the Standard way. If CostLog is defined you will have to change the cost settings of your nodelist compiler. BT EE needs the cost in tenth of seconds for one cost unit in the compiled nodelist. You will have to redesign your nodelist compiler configuration. Here's an excerpt from PARSELST.CFG: Cost 386 44 ; ^ ^ ; ! übriges Ausland: 4.42 Sekunden ; übriges Inland: 38.57 Sekunden ; wegen Speed die USA zuerst: 1- 44 ; Nahbereich (720 Sekunden): 49-6451 7200 49-6452 7200 49-6457 7200 ; ... ; Zone 1 (120 Sekunden): 49-2732 1200 49-2725 1200 49-2733 1200 ; ... ; Ausland 1 33- 160 ; Andorra 32- 160 ; Belgien 45- 160 ; Dänemark 298- 160 ; Faröer 33- 160 ; Frankreich 30- 160 ; Griechenland (ab 2000 Uhr) 44- 160 ; Großbritannien & Nordirland 353- 160 ; Irland 39- 160 ; Italien (ab 2000 Uhr) ; Lichtenstein 352- 160 ; Luxemburg ; Monaco 31- 160 ; Niederlande 43- 160 ; Österreich 351- 160 ; Portugal (ab 2000 Uhr) ; San Marino (ab 2000 Uhr) 41- 160 ; Schweiz 34- 160 ; Spanien (ab 2000 Uhr) 42- 160 ; Tschechoslowakai ; Vatikanstadt (ab 2000 Uhr) ; Ausland 2 20- 106 ; Ägypten 213- 106 ; Algerien 359- 106 ; Bulgarien 358- 106 ; Finnland 350- 106 ; Gibraltar 354- 106 ; Island 972- 106 ; Israel 962- 106 ; Jordanien 38- 106 ; Jugoslawien 218- 106 ; Libyen 356- 106 ; Malta 212- 106 ; Marokko 47- 106 ; Norwegen 48- 106 ; Polen 40- 106 ; Rumänien 46- 106 ; Schweden 963- 106 ; Syrien 90- 106 ; Türkei 216- 106 ; Tunesien 7- 106 ; UdSSR 36- 106 ; Ungarn 357- 106 ; Zypern End [EuroCost] - Special gimmick for "Hanstheo Wolf @ 2:242/4": If you define NoWildCards "NoWildCards" then wildcards are not allowed in the filename part of a request. This doesn't concern the usage of Wildcards in the extension, though (when you doesn't know if the archive is LZH or ZIP). Somebody used to request A*.* from Hanstheo ... [NoWildCards] - the costlog (for national purposes) has become two formats. One "big" if abs(CostLog) is 5 and one "small". [EuroCost] - define the cost unit with "CostUnit n". If you leave it out, CostUnit this Binkley uses 23 (for germany every n seconds cost 23 Pfg, n differs with the distance of your call). If you live in switzerland you will probably define "CostLog 10" for 10 Rappen. [EuroCost] - The behavior of ScreenBlank was modified. The screen will be lit again only when a call becomes a CONNECT, not when a call will be initiated. [ScreenBlank] - During zmodem-filetransfers Binkley will remember the last blocksize and use this blocksize as startvalue for the next file to be transfered. [ZModemBlocklen] - An empty line on program startup isn't no longer emitted to the logfile. [Space] - If you call a host you will now no longer get the "Called xxx/0 and got xxxx/1" message. [NoFlame] - If you do with ALT-S or ALT-G a request then binkley add a comment into the file that he did made this request. But the other binkley complaint about this, saying that he couldn't recognice this. So this feature is stripped. [NoComment] - Who needs a SuckerLimit? Mark (using bonk) this fellow in the SuckerLimit nodelist with one of the unused bits. Keep the bit in mind, for example Bit 10. Now write an "SuckerLimit A 10" in your BINKLEY.CFG ('A' is 10 in hexadecimal). And watch, what happen after 10 minutes of file requests by the sucker. [Sucker] - Original BinkleyTerm sometimes write "Nobody answered the door", this is due to an error in the state-table (which i haven't found, sorry). But now Binkley doesn't hang up after this message instead he is trying again. Most times the session will then start! [NoBody] - The "SmallWindow" in BINKLEY.CFG now has an additional effect: if defined AND if the current session is not reliable (no MNP-link) then transferring data via WaZoo will begin with small blocks. Nice for bad lines, because there is no need to resend 2048 bytes to correct a single error. The additional overhead due to the smaller blocksize is not as big as you might think. This is not the case when you have an /ARQ-session. A remembered blocklen (from the file lastly transfered) has priority over this scheme. * If it's an ARQ connect, Binkley will step up the block size as fast as possible. [ZModemBlockLen] - some seconds after start of binkley you now see who all has modified this binkley. [Version] - The text "Modem protocol negotiation filtered" now wont be written to the screen and in the logfile. This text is meaningless to a normal user. The logic of the mnp-negotiation- filter isnt changed, however. [DontWorry] - Changed the format of the "Pending Outbound Mail" Window: NiceOutbound The "Node" field now is only 12 characters long. This is usually enough. New: The "Flags" field. It carries the information spread about half the window in the standard version. Depending on the files in the outbound directory the characters C/H/D/N and/or R are displayed (for crash, hold, direct, normal and requests). The old "S" field (status) has been removed. Instead, a human readable text is displayed if an error occurs: Unknwn for nodes that have no entry in the nodelist NoConn when the maximum values in BINKLEY.EVT ("T=n,m") are exceeded. Tried If a session took place, but some mail or requests for the node still remains. New: The "Files" field displays how many files are pending for the node. All *.?UT, *.?EQ, *.?LO and all files in flow-file(s) are counted. New: "Age" field displays how old the oldest of all the files to be sent is. Nice for people with points not picking up their mail regularily. The date of the files inside of the *.?LOs flow-files are not considered, because you can send very old files from your system via file-attach. ┌ Pending Outbound Mail ────────┐ │Node Files Size Age Status│ │1:135/20 1 267b 0 C│ <- a crash to joaquim │2:243/1 1 7.1k 0 NR│ <- normal mail and a file request │2:242/29 2 127k 1 H-│ <- 2 files since yesterday on hold │ │ └───────────────────────────────┘ - There's no wrap at the end of pending outbound list anymore. * PGUP, PGDN an END now work correct in the Pending window (bugs in original 2.5 release). - If you define "NiceOutbound" in BINKLEY.CFG, my BT version NiceOutbound reads in all additional information. This can take a moment, but it's worth it... If you haven't defined it, the behaviour is the same as in the normal release. - The "Pending Outbound" window is now nicely sorted and it doesn't wrap around anymore. [Sorted] - You can toggle between few informations (fast) and many infos ALT-F with ALT-F. [NiceOutbound] - When BinkleyTerm reads in the outbound directories, it now shows "Reading" in the Settings window. [WhatIDo] - No BINKLEY.DAY is written anymore [NoDay] - Correctly process disabling of WaZoo pickups. [NoWazooPickup] - New DOS-Shell procedure [Exec] - Requesthandling has changed in some places: - Don't rename incoming REQ Files [DontRenameReq] - Make no FLO file for requests [NoReqFlo] - Try to request any time [TryReq] - Don't call for requests if not allowed [Request] * In Send and Get Files is now possible to enter more than ALT-S one file name. Enter filename, password ("!") and other ALT-G features each on a line. There are three lines per popup window. If you entered three files than BT will ask you after the "Crash/Hold" about "More (Y/N)". If you type <Y> you'll be able to enter even more files. The More-Prompt does appear until you enter less than three files. [OneMore] - If you dare to type ALT-Z, you will be surprised seeing a ALT-Z very large, new window! It's the so called "ZOOM WINDOW". It holds detailed informations about all the things that are waiting in the outbound. Furthermore, this window has it's own online help. You can have a look at it typing ALT-F10. ZOOM shows Node, Files, Bytes, Age, Calls, Bad Calls, Status and Errors. If there's no action for some time or there's action from the modem (e.g. RING) the ZOOM window closes and returns control to unattended mode. ALT-Z only works if the video fossil is loaded (DOS)! There's no check if it really is loaded. But I think this is no problem because there's no sense in not loading the video fossil which speeds up BT's video output! [Zoom] * I expanded the ZOOM window with the most important outbound handling routines. There's now a scroll bar which can be moved with the UP, DOWN, PGUP, PGDN, HOME and END cursor keys. So you can choose a specific node. Following new features are implemented: - ALT-F10 Help screen with ZOOM functions - A Readdress Mail - G Get (request) Files - I Reset Dial Tries - P Poll any System (Create CLO File) - C/D/H/N Change to Mail to Crash/Direct/Normal/Hold - R Delete Requests - S Send Files - T Stop Calls [Zoom] * NiceOutbound and "Zoom Window" have been adapted to BT 2.5 internal functions. * There's now a Task Number (if defined) in the "Current Setting" Window wird. [TaskNumber] * ReInitTime now can be defined up to 60 minutes, the old max. ReinitTime setting has been 15. Default is 10. * New CFG Token: PktRsp PktRsp PktRsp tells BT to send an answer in PKT style instead of a RSP file. PktRsr only works when a FlagDir is defined, because it creates the PKTs in this directory. The PKTs will be deleted at beginning of the next session. They are coded with the task number, so that no crashes will occur! I've changed the behaviour of these files many times and I think this solution now is nearly perfect! [PktRsp] * OKFile extension: MaxAreas Adding FileSec *path\MAXFILES.IDX KnownSec in OKFILES.TXT lets BT search through this database when ProtSec processing file requests. This is much faster than seeking through whole a partition (and it's more "multitasking friendly", too!). You will have to add following keywords to your BINKLEY.CFG: FileSec n KnownSec n (optional) ProtSec n (optional) where n is the following security level: 0=Disgrace, 1=Limited, 2=Normal, 3=Worthy, 4=Privil, 5=Favored, 6=Extra, 7=Clerk, 8=Asstsysop, 10=Sysop, 11=Hidden, -2=Twit (not supported) Following line has to be added to your BINKLEY.CFG, too: MaxAreas d:\max\area.dat Each caller will be assigned a "user level" (security) depending on their caller level (default, known, prot). This security will be compared to the access rights in the Maximus file area. If the security is too low, the caller will get a password error. Access to 5000-8000 files is very fast: it takes about _one_ second on my setup! This won't be very much longer on slower systems because the file access time will stay about the same. There will appear some new entries in the log with details to each search if you have LogLevel +/-5 defined in yout setup. I got many nice comments about this new features. Some people have been very happy about the fast file finds even with wildcards! [MaxFiles] * Shells are now possible everywhere. - Shells are generated with negative errorlevels (-1 to -9) - Shells sometimes don't work together with dynamic events So please use the "Aftermail" feature of BT 2.50! Maybe Vince can tell me what's going wrong sometimes! [FastEvent] * Session Statistics: BT EE display now during a session the "Current Session Statistics": - Sent, Rcvd: all data calculated only over real transfer time. - Total: Data is calculated over the whole session time (since the first file transfer). - Wait: Data from the "Pending Outbound"/"Zoom" window - There is no "Wait" data without NiceOutbound. - BT adds the SSTat data to the CostLog if LogLevel is +/-5. The SStat window is not overwritten until the next outbound scan or keypress. [SStat] * File names will now be searched very soon for path chars as ":" and "\". Matching files are deleted instantly. * Flags (which are scanned in the flags directory): - BTEXITSx.## creates shell x on task ## (Hex) - BTEXITxx.## creates errorlevel exit xx (in Hex!) on task ## (Hex) - BTRESCAN.## makes BT rescan the "Pending Outbound" on task ## (Hex) [ExitFlag] * The whole flag handling is now dynamic with up to MAX_AKAs flags... If you activate "MakeDir" (see below), missing PNT directories will be created autmagically. The flags will be created in the flag directory only if there's no other solution! This makes the flag handling capable of a good EMSI implementation with multiple aka support! [Flag] * New CFG Token: MakeDir MakeDir When "TaskNumber" is active and if BT creates flag files, not available directories (PNT, zones) are now created on the fly. Empty PNT directories are deleted after use. [Flag] * Help screen as Popup Window (needs BINKLEY.LNG changes!) [NewHelp] * All Helps and other displays are now with timeouts. Also when receiving a char from the modem ("RING") BT returns to unattended mode. * strings are cut after a ";" when reading BINKLEY.CFG * New CFG token: NoArq arqstr NoArq - repeated up to five times in BINKLEY.CFG - you can define here alle extended connect string which should'nt passed to the bbs. If none of these strings matches, BT displays "ARQ Connect" in the "Current Settings" window - My setting for a HST-DS: NoArq /None NoArq /V32/None [NoArq] * About all texts and keys are defined in BINKLEY.LNG, so that there will be nothing to stop your creativity. I'm afraid that no key can be defined twice in Unattended and Zoom mode. So all keys defined in Zoom mode don't need <alt> to be pressed. * Pressing <end> in an empty outbound produced a bad trap. END Fixed. * If you activated "PrivateNet", all your points are now displayed correctly in 4D. Your own points can now be treated everywhere as 4D even if BT works in 3D (privatenet) mode. [4D] * CostLog is now updated in Janus mode, too! * Fixed bad structure bug in file_dos.c (wr_time and wr_date are in wrong order!) * I don't like that serial number gimmick. [NoSerial] NEW FEATURES IN BETA-B ---------------------- Tobias Burchardt and me exchanged our source code changes, so I was able to implement EMSI into my BT version without having to reevent EMSI the third time... I updated his sources in _many_ aspects and it will be improved even more in the next beta releases. There still are some nice ideas... This is a fully featured beta-version of BinkleyTerm that supports the EMSI-(FSC-0056)-handshaking method. I suggest you to enable to "MakeDir" and disable the "PrivateNet" keyword in BINKLEY.CFG to let EMSI work reliable. * How does my EMSI implementation react to: * wrong password(s): - immediate hangup - this "feature" is still under discussion. maybe I will make it configurable in a future version so that all akas with wrong password will be simply ignored. * up to three akas: - create all BSY flags if possible * more than three akas: - only create those akas for which an outbound directory exists. This will speed up session handshaking and normally won't cause problems on multiline/multitasking systems! * aka is busy: - don't send to that aka * main address (called address) is busy: - hangup immediately [EMSI] * Due to the possiblilty of EMSI to transfer site-information MyLocation some new keywords have been added to BINKLEY.CFG: MyPhone MyListFlags * MyLocation - location of your system MyMaxBaud * MyPhone - your phonenumber (data) * MyListFlags - your nodelist-flags * MyMaxBaud - the max. baudrate supported by your modem (300-9600) - Examples: MyLocation Emmenbruecke CH MyPhone +41-41-538607 MyFlags XA,V32B,HST,V42B MyMaxBaud 9600 - As you see, that these info can be taken from your nodelist! * NoEMSI [flag] NoEMSI - NoEMSI without [flag]: - disable EMSI-support for callers and outdials - NoEMSI with [flag]: - [flag] can be "A" trough "F" - if the same modem flag (in the nodelist!) for the dialed system is set, BT won't use EMSI in this session! I still don't know why all the people want EMSI in their mailers. I tend to use for all "reliable" and often polled Nodes the NoEMSI feature to poll with WaZOO which is faster and more reliable on bad lines. * PickUpAll PickUpAll - when connected via EMSI, pickup mail for ALL akas. If not specified, you'll only get mail for your primary aka (beware of BT's automatic aka-matching!) * ExitBaud [<errorlevel>] <Connect> ExitBaud - This is a new feature for people who have to react on different connect messages (e.g. Fax modems). - If BT encounters the given <Connect> messages reported by the modem it takes the following action: errorlevel > 0: * forces Binkley to exit with the specified errorlevel (or 99 if not defined) * it generates a batchfile called BAUDBAT.BAT that contains the line "EXITBAUD <locked Baudrate> <Baudrate> <Port/Handle> <Time_to_next_event> <Connect>" - errorlevel < 0: * forces Binkley to shell to call "EXITBAUD <locked Baudrate> <Baudrate> <Port/Handle> <Time_to_next_event> <Connect>" - For example you have a Fax-Modem that reports 'CONNECT FAX' and you want to call your fax-software as fast as possible, you may include ExitBaud 199 FAX into your CFG. You see that this statements works like the ExtrnMail-possibility... - In OS/2 you'll have to use the shell! [BaudExit] * NoMNPFilter NoMNPFilter - This statement disables the MNP-filter and causes Session-handshakes to start faster. Us this with new modems which support MNP/V.42b; they suppress mnp/v42 garbage well enough. * MinBaud, KnownMinBaud, ProtMinBaud MinBaud - these 3 keywords define the lowest baudrate at which KnownMinBaud frequesting is allowed. If the caller is at a lower ProtMinBaud baudrate he'll get the right response-message. Have a look in SAMPLE.TPL for the new 'status 8'! [MinBaud] * Again new flag handling. If you define "MakeDir" (recommended for EMSI sessions) BT will now create _all_ outbound directories automagically. It will delete mpty .PNT dirs but it won't delete empty zone outbounds! There will be no detailed information ofcreating and deleting BSY flags during EMSI sessions (only errors and a summary are logged). * Shells with dynamic events should now really work! A BREAK THROUGH until sb tells me the contrary. :-)) Some temporary version of this beta still had some problems. I had to think over the whole thing again and now I've found a very nice solution. Don't be scared if BT-EE sometimes doesn't rescan the outbound after an emsi session! [FastEvent, EMSI] * I updated ENGLISH.TXT once again: - if you want to have the processing of EMSI akas logged, change the ">Processing..." line to " Processing:...". - Change back ">No BOSS in nodelist" if you need. - Some more additional texts * TABs in the .CFG files will now be converted to spaces. * Incoming REQ files are now tanslated from REQ to R## (where ## is your task number in hex) if you've defined a task number. This gives you the possibility to have one inbound directory with no interfering REQuests! [Req] * Added status 7 = security level too low (maxfiles) 8 = baud rate too low (minbaud) in my SAMPLE.TPL * If you're running un fullscreen mode and enter ALT-M BT-EE ALT-M will ask you now for "Fast or Crash", too. If you enter "F" for "Fast", BT will poll like crazy (compatible to BT 2.5's behaviour ;-)). If you enter "C", BT-EE will create an empty CLO file for the polled node. I removed the "Y" hidden feature again (sorry, Alex). [Poll] * CFG-Errors will now be written to stderr instead of stdout. When running OS/2, you're now able to redirect all CFG errors to a special file using stderr redirection, e.g. "BTP .. 2>BINKLEY.ERR". I don't know if DOS supports such nice things, too. ;-) [CfgErr] * Updated "ExtrnMail" to work with Shells. ExtrnMail [FastEvent] * Fixed a bug in the readdress function in ZOOM mode. It was impossible to readdress mail to another zone. NEW FEATURES IN BETA-C [8.03.92] --------------------------------- * fixed some nasty bugs in the EMSI implementation. Memory wasn't initialized ok sometimes and it was possible to cause lockups when large emsi blocks are transferred. * Entering ".X" as address caused some misbehaviour in ZOOM readdressing. * Redesigned again the alt-M poll feature. It should now please everybody... * Alex is reporting problems when reading in large outbounds with long v6 nodelists. Did anybody else encounter the same problems? * Uups: Level 7 "security error" in sample tpl was used since BT 2.5! So it's Level 3 again. NEW FEATURES IN BETA-D [11.07.92] ---------------------------------- After some time of inactivity, I'm back again with some IMHO neat bug fixes and additions. * BT sould now delete all outbound directories it had to make when creating BSY flags. It does NOT delete directories which existed before. [Flags] * Redesigned screen layout so that the pending outbound window is wider (5D addresses look nice now!). No more unused space! [MB] * Tried to implement some of Chris Laforets ideas of multithreaded communications. But I didn't see any improvements. It took more resources than the original version. MAXCOMM.DLL would be a nice solution, but I don't want to reinvent the wheel. * Implemted 5D support into the EMSI code. [Emsi] * Changed some code concerning reentries after shells. [MB] * Task number is now located at upper left corner [MB] * Fixed traps when receiving requests which consist of various wildcards (e.g. "*AB**?.*"). [MB] * Hopefully fixed some other trap situations with password protected file requests. [MB] * Hint: Do you know that it's possible to enter sysop names instead of node addresses to choose a node? * All node addresses entered in an outbound management function appear with their BBS name if available. If the node is not found in the nodelist(s), no special info is shown. [MB] * Cursor is now visible while typing at an input prompt. [MB] * "Still have mail..." now doesn't appear after BBS shells. [Emsi] * Killed V6 support but reenabled it again after protests. :-) It would save about 2kB of EXE size! * ExternMail should now be available with 8 bit support. [MB] * Terminal mode: after dialouts there's been no TermInit again: fixed. [MB] * OS/2 version supports now up to 57600bps locked ports. [MB] * EMSI AKA checks: LazyAkaChk - There's been some suggestions with bad passwords on EMSI Akas. Standard behaviour is to hang up if the remote password does not match our passwords. If you enable "LazyAkaChk" in BINKLEY.CFG, all Akas with bad passwords are deleted (no mail is sent to those akas) but BT does NOT hang up. This feature is very useful at various situations. The suggestion came from our host who experienced problems with changed akas of his up- and downlinks. - If you want to protect _your_ akas, e.g. tell BT to hang up if a remote system calls you with one of your akas in his aka list, install a password for your own akas! This "feature" can be combined with "LazyAkaChk", too... [Emsi] * Now to something very special: FAX RECEIVE! FaxIn ************************************************************ IMPORTANT: Fax receive ist really an ALPHA version and I only know that it works here on MY system. So if it does not work with your setup, don't be too diappointed. If you think that there's only one or two little patches missing which will make it work, please turn on debug mode and send me the relevant parts of your log file. As I don't have two faxes at the moment (and not to mention a small indoor phone exchange), it's a little bit hard (and expensive to test it under real conditions. ************************************************************ I took me some time to implement this feature because there are no fax docs available. I tried to convert some UNIX sources but they didn't work. Well, after all I got the CCITT draft sometime and after a short time I was able to receive faxes with my ZyXEL modem. Because I don't have any other modem than the a ZyXEL I don't know if it works with other modems, too. If they meet the CCITT proposals it should work, because BT receives faxes totally CCITT compliant (well, with some additions to be even more "compatible"...). How to enable class 2 fax receive: - Lock your port to 19200 ("Baud 19200") - Init your modem to class 2 operation (NO proprietary fax are supported), maybe send your local id to the modem - Include "FaxIn D:\Fax_Inbound_Path\" to your BINKLEY.CFG - Wait for a fax transmission. :-) - Each fax page is stored in the FaxIn directory as PAGExxyy.FAX. (xx = task number, yy = page number). The file format is pure G3 FAX. You should be able to convert it to PCX, TIFF etc. with some freely available conversion programs. Examples: - How to setup your ZyXEL modem: * switch you modem off and on * go to terminal mode and type AT #B0+FCLASS=0 (which turns off ZyXEL string and switches to data/fax modem operation) * edit "Init" in your BINKLEY.CFG: e.g. ATZ[.....]+FLID=<id-string> (Zyxel does only accept numbers as id-string although it should be possible to enter 20 ASCII characters) (ATZ is very important because the ZyXEL firmware does not fully reset the modem after a fax connect). * wait for a fax... Known Problems: - Some fax programs do not correctly initiate more than one page/document. I don't have a work around for this bad behaviour yet. If you encounter any problems, please turn on Debug mode in BINKLEY.CFG and send me some log file extracts. - Some older EPROM revisions of ZyXEL modem do not raise carrier until phase C data transmisssion. To be able to abort a bad fax connect fast enough, BT should check CARRIER before phase C. Since Zyxel implemented this feature very late in R4.09, I had to disable the carrier checks at the beginning of a fax connect. This makes BT wait up to one minute for a proper fax connection. I'll have to change that in the near future. - It's a pity that I cannot lock higher than 19200 in fax receive mode. I'm working on a solution to this problem (maybe raise rate on normal connects: I don't know yet if that works.). Any comments and hints are welcome. This feature does not disable external fax receives! [Fax] * "NewNodelist" and "QuickNodelist" are not longer valid tokens (replaced by "LazyAkaChk" and "FaxIn"). [Emsi, Fax] * Disabled debug infos in the state machine to make it faster. [MB] SUMMARY OF NEW FEATURES AND BUGFIXES IN BETA-E [12.09.92] ---------------------------------------------- * 32Bit under OS/2 using IBM C Set/2 compiler * Beta-E (OS/2) uses now MAXCOMM.DLL * Send/Receive buffer of MAXCOMM.DLL can be adjusted as before: SET TBUF=... SET RBUF=... I use T/R 512/4096 on my 386/40 * The following text may be of use for all MAXCOMM users: When using COM0x.SYS, the default buffer size is the same as the buffer size used by COM0x.SYS. For Tx, this is the optimum size to keep COM0x.SYS busy. A buffer larger than this will not gain any extra performance. Therefore, the common urge to make the transmit buffer large will not have the desired effect. If there was some way to make the REAL transmit buffer (the one in COM0x.SYS) larger, performance gains WOULD be seen, but there is no way to do this unless you have the source code for COM0x.SYS. If you DO have the source (in the DDK), and you increase the buffer sizes, use 0 for the XxBufSize and ComOpen will use your larger buffers (ComOpen/Register uses the standard IOCTL call to query the buffer sizes). RxBufSize, on the otherhand, does not affect performance of any functions in the Com() module. To avoid overflows however, you should set this to at least the size of the COM0x.SYS Rx buffer, or use 0 so the default size if selected. When the Rx buffer gets full, the speaker will sound (DosBeep(200, 10)). There are two solutions to this problem: Make the Rx buffer larger (preferred), or raise the priority of the thread calling ComRead()/ComGetc(). All threads created by comm.dll run at the default priority. Since com0?.sys is interrupt driven, Reading/Writing to it at a high priorty does not improve anything. If the Rx thread is starved for such a long time that com0?.sys buffer gets full, it means some other process on your computer is badly behaved (IT is problably running at Time Critical class when it shouldn't be). There are four ways to solve this (in order of preference): 1) Insert DosSleep(1L) in the polling loops that are guilty (this assumes you have source code), and/or have it run at NORMAL priority class. 2) decrease MAXWAIT in config.sys. 3) Prior to calling ComOpen/ComHRegister, raise the priorty of the current thread (or of the whole process). The Rx/Tx threads will inherhit this priority. 4) set PRIORITY=ABSOLUTE in config.sys. FYI: OS/2 1.2 uses a default size of 1024 bytes for Rx, and 128 bytes for Tx. * MAXFILES.IDX search should now work with all file names. * New token: "CheckInit" BT watches now the modem reponse after issuing the init string. If there's no "OK" coming, BT tries again. After two unsuccesfull tries, BT exits with errorlevel 2. I recommend to comment out "PreInit" when using this feature. * "Share" is now default in the OS/2 versions. * EMSI "phantom mailer" bug fixed. * "Double outbound" won't be scanned double. * Spaces in front of comments (";") in *.CFG are now deleted. * Larger stack size * Simple domain matching on all domain-less nodenumbers if you're using a 5D setup. See BT_REF.DOC for details. * New product codes in ENGLISH.TXT (thanks, Vince ;-)) * Externel BINKLEY.HLP. It is read in the current dir or at the BINKLEY environment path (same procedure as BINKLEY.LNG). This file is built up as ENGLISH.TXT. The first char in each line shows to which help screen the following line belongs: U = Unattended Mode Help Z = Zoom Mode Help T = Terminal Mode Help The first line with "U" and "Z" is used as title for the popup window. See the attached BINKLEY.HLP. BT stops after each page. Leave help pressing ESC or continue pressing any other key. * Event flag "p" substitutes NoPickup in BINKLEY.CFG * FD-style REQ files at point systems should now be processed right. * Class 2 fax receive works again. Many thnaks to Matthias Hertzog, who sent me a fax modem for testing purposes. * Just enter .pointnumber when addressing your own points. * "NiceOutbound S" sorts the Outbound as usual, "NiceOutbound" sorts as the "Original" :-) * Alt-K (Kill) works again with unknown systems * Time to next Event now in hours/minutes * Wrap at system display in popup windows fixed. This has been the reason for some bad traps and misbehaviours. * ModemTrans works now as in BT 2.56. It'l check 1:1. * Roland implemented his input routine and I couldn't resist changing the whole thing again... * Smaller EXE size due to putting togheter some code fragments. * Two new functions: Alt-N: Toggle NoPickup Alt-U: Toggle Send only ATTENTION: If you toggle those two flags they are saved PERMANENTLY in BINKLEY.SCD! So if the same event reappears one day later, it'l have the same status as the day before! * Long file name support in the 32bit OS/2 version on HPFS partitions. Filenames used in XModem/Sealink send are truncated. * Chnage in OKFILE list for the 32bit OS/2 version: It has been possible to use various path and filename/wildcard combinations. This has been changed now. All entries with "D:\path\FILE*.ZIP" or similar wildcards/filenames are not longer valid. The only valid path entries are now "D:\path\" or "D:\PATH\*.*"! This does NOT affect entries with MAGIC filenames! This change has been necessary to allow the support of long filenames. As a side effect, the directory searches are now VERY fast. Have fun! Michael Buenter, 2:301/602