ftp.update.uu.se

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

/ ftp.update.uu.se / ftp.update.uu.se.2014.03.zip / ftp.update.uu.se / pub / rainbow / msdos / decus / RB141 / varug04a.arj / V4N2ASC.ASC < prev next >

Wrap

Text File | 1990-03-19 | 66KB | 1,117 lines

Vancouver Area Rainbow Users Group N e w s l e t t e r March and April, 1990; Volume 4, Number 2 Editor: David P. Maroun, 9395 Windsor Street, Chilliwack, BC, Canada V2P 6C5; telephone (604) 792-4071 Publisher: DECUS Canada, 505 University Avenue, 15th Floor, Toronto, Ontario, Canada M5G 2H2; telephone (416) 597-3437 ----------------------------------------------------------------------------------- Unless the contrary is indicated, any part of this newsletter may be freely copied or distributed unaltered and with credit given to the original source. While the information provided is believed accurate, the editor cannot take responsibility for contributions of other writers. ----------------------------------------------------------------------------------- Deadlines: For our May and June issue: April 30, 1990 For our July and August issue: June 30, 1990 Almost any legible format is acceptable for submissions, but the ideal is ASCII form on diskette. Diskettes should be accompanied by covering letters describing the files and indicating disk format. We can handle Rainbow CP/M, Rainbow MS-DOS, IBM PC-DOS single-sided, and some others as well (check about them). ----------------------------------------------------------------------------------- Editorial: A Good Follow-Up by David P. Maroun During DECUS' leadership meeting in Ontario last Fall, I shot two rolls of color slide film. I gave copies of several dozen pictures to the DECUS office, and some were used in the Winter 1989 issue of DECUSCOPE. In fact, every picture in that issue, except the photo of Mark Johnson on page 2, was originally mine. The final pictures were black-and-white prints reproduced for publication, so several steps were necessary to process my color slides. Nevertheless, the results were technically more than acceptable. So, here is a compliment to the DECUSCOPE editor and those who processed the pictures. They started with originals which were not ideal, but still got good results. ----------------------------------------------------------------------------------- No More Prices DECUS has updated its policy on commercialism. To agree with that policy, this newsletter will omit prices for merchandise it describes or lists. ----------------------------------------------------------------------------------- Corrections And Clarifications An article in a previous issue of this newsletter recommended G. Allen Morris III's Disk Organizer (DOG), a shareware file defragmenter for MS-DOS. Some recent trials indicate possible problems in using this program on a Univation 11-megabyte non- booting hard disk, such as was installed in some Rainbows. When run on one of these disks, DOG reported that the two copies of the file allocation table did not match each other. In one case, DOG was told to go ahead anyway, and some files were later found cross-linked. So, we advise caution before running Disk Organizer on Univation disks. We suggest you back up any disk before running any file defragmenter on it. ----------------------------------------------------------------------------------- Charlie's Museum Of MS-DOS Horrors: The End (?) by Charlie Gibbs (Editor's Note: This article originally appeared in the West Coast Computer Society's "WCCS Printout".) Both MS-DOS and the processors on which it runs are steeped in tradition. The 8088, 80286, and so on are direct descendants of the 8080, while MS-DOS is a descendant of CP/M. Both hardware and software pay homage to their ancestors in many ways; one of the more perverse ways is by inheriting limitations which are no longer necessary. A hardware example is the 8080's 64 k address space, which encumbers 8088 programmers with messy segment calculations. But here I'm going to focus on a software quirk which has outlived its usefulness. I'm referring to the single byte found at the end of most text files. It contains a hexadecimal 1A (CHR$(26) for you BASIC programmers), and is used to denote the end of data in a text file. The CP/M file system stores the length of a file as a number of sectors rather than bytes. The length of a CP/M file can only be determined to the next multiple of 128 bytes. Text files, though, can be of any length, and additional garbage at the end of the file is intolerable. Therefore the designers of CP/M set aside the hex 1A character as an end-of-file marker--a 'magic cookie' on which all text readers would stop. This trick wasn't necessary in MS-DOS, whose file system stores the length of a file right down to the byte. But Microsoft, bowing to tradition, decided to leave in the EOF marker anyway. This extra byte has probably caused me more grief during software development than any other single quirk of MS-DOS. First of all, is this byte really a part of the file? Let's find out. Bring up GWBASIC and type in the following program: 10 OPEN "O",#1,"TEST.DAT" 20 PRINT #1,"The quick brown fox jumps over the lazy dog." 30 CLOSE #1 40 OPEN "I",#1,"TEST.DAT" 50 PRINT LOF(1) 60 CLOSE #1 70 OPEN "A",#1,"TEST.DAT" 80 PRINT LOF(1) 90 CLOSE #1 100 OPEN "R",#1,"TEST.DAT",1 110 FIELD #1,1 AS I$ 120 PRINT LOF(1) 130 CLOSE #1 140 END Now RUN it. The program will print: 47 47 47 Exit the BASIC interpreter (by typing SYSTEM). Once back at the DOS prompt, type: DIR TEST.DAT This will show that the file is indeed 47 bytes long. If you have some sort of hex dump utility (such as PCTools or Norton), take a look at the file. You'll see that the 47th byte is a hex 1A. The byte definitely seems to be part of the file, although any program which reads the file sequentially will get an "end of file" indication as soon as it sees the hex 1A. When the above program is compiled (using the IBM BASIC Compiler version 2.00 [by Microsoft]) and the resulting executable file is run, the following results are printed: 47 46 47 (Editor's note: I got the same result with the DEC Rainbow GW-BASIC compiler version 2.0.46 as Charlie Gibbs got with the IBM BASIC Compiler 2.00.) BASCOM doesn't count the hex 1A in the file length if (and only if) the file is opened for append. Note that this is inconsistent not only with GWBASIC, but with other file modes within BASCOM itself! This can cause all sorts of fun when moving between the interpreter and the compiler. File size calculations can be a nightmare. Now type the following command: DIR >> TEST.DAT TEST.DAT This should append a directory listing of our test file to the test file itself. Type: DIR TEST.DAT and notice that the file has indeed grown. But if you try to TYPE TEST.DAT all you'll see is "The quick brown fox jumps over the lazy dog." If you take a hex dump of the file, though, you'll see that the directory listing is actually there. It's been placed behind the hex 1A character, which stopped the TYPE command from reaching it. If you wrote a BASIC program which opened the file for append, the new text would overwrite the original hex 1A byte and continue from there. Console redirection, being used for text, should be smart enough to do the same. But hey, "that's not a bug, it's a feature!" So beware of the dreaded hex 1A. It's poison to text readers. If you're writing a program to capture raw text from the serial port into a file, be sure to take the trouble to filter out any hex 1A bytes which you may encounter. Otherwise, a burst of line noise might corrupt an incoming byte into an end-of-file marker, and when you try to TYPE the file or read it sequentially by a program, you'll lose everything after the garbled portion. I don't know whether Microsoft finally eliminated the hex 1A byte in OS/2, but given their past record I'm not too hopeful. Maybe someday I can repost these columns as "Charlie's Museum of OS/2 Horrors." (Editor's note: Some editors--recent versions of SEDT, for example--permit the hexadecimal 1A character [also called 'control-Z'] in a file. A final control-Z is required on a text file for some MS-DOS applications. XMODEM transfer is one example. A text file also needs a final hexadecimal 1A before being transferred to CP/M. Omitting the end-of-file marker in these cases will result in visible garbage being appended. The MS-DOS 'TYPE' command will not read past a 1A, but you can see everything in TEST.DAT by entering 'COPY/B TEST.DAT CON'.) ----------------------------------------------------------------------------------- Sources Of Used DEC Equipment Here are some parties who recently offered used products originally made by Digital Equipment Corporation: United Products Incorporated 1123 Valley St. (right by the Mercer St. exit) Seattle, Washington Telephone (206) 682-5025 Jim Greely Telephone (617) 769-9811 (located in Massachusetts) In some areas, call toll-free (800) 446-3074 Que Computer Systems (ask for Steve Laurie) 7100-44 Street SE Calgary, Alberta T2H 1X2 Telephone (403) 236-6100 United Products and Jim Greely have used Rainbows. Que Computer Systems has used LA50 and LQ502 printers. ----------------------------------------------------------------------------------- Using A Color Computer With An LA100 Or LA50 Printer by David P. Maroun I often use a Tandy Color Computer 3 for writing letters, keeping track of my personal finances, and other mundane tasks. I have usually connected the CoCo (as a Color Computer is called) to a Tandy printer which is designed for that computer, but recently I decided to use a Digital Equipment Corporation LA100 and LA50 instead. The CoCo and all three printers have RS232C serial ports, but the Color Computer and Tandy printer use 4-pin female DIN plugs while the DEC printers have 25-pin D-subminiature male connectors. So, while Radio Shack sells cables for connecting the computer to the Tandy printer, I had to make up a cable for the DEC machines. Materials I needed were: A male 4-pin DIN connector, three-wire cable (I used four-wire telephone cable and ignored the fourth wire), a 25-pin female D- subminiature connector, a soldering iron, and solder. I got both connectors from Radio Shack. Hindsight tells me I should have looked elsewhere for the D- subminiature. That connector's bolts are too short to reach the thread holes on the DEC printers, so only weak friction maintains contact. I made the following connections: ---------------------------------------------- | DIN plug | D-subminiature | ---------------------------------------------- | Pin 2 | Pin 11 | | Pin 3 | Pin 7 | | Pin 4 | Pin 3 | ---------------------------------------------- The computer sends information from its DIN plug's pin 4 to pin 3 of the printer's D-subminiature; pin 3 of the DIN plug and 7 of the D-subminiature are grounds; and the printer uses its pin 11 for sending a busy signal to pin 2 of the computer when the printer buffer is full. The DEC printers can use XON-XOFF or ready-busy signals to control the flow of information. The CoCo uses ready-busy signals, and I configured the printers accordingly. On the LA100, I set dip switch B1 to the right. I also had switch B2 to the right for no modem control, but I am not sure whether that setting made any difference. On the LA50, I closed SW 1-6 to enable ready-busy protocol, and also closed SW 1-7 to make busy be low and ready be high. The LA100 was set up for transmission at 9600 bits per second. For this purpose, switches A1, A3, and A4 were at the left; A2 and A5 at the right. I enabled this transmission rate on the Color Computer with the BASIC command POKE 150,1<ENTER> or its equivalent in the Telewriter-128 word processor I use. On the LA50, I used the factory-set (and highest) transmission rate of 4800 bits per second. This required switches SW 2-1, SW 2-2, and SW 2-3 to be open. On the CoCo, I got 4800 bits per second by the BASIC command POKE 150,7<ENTER>. The Telewriter-128 manual says to use 6 instead of 7. When I tried 6, I got a garbage output. Text was printed as intended when I poked a 7 into location 150. The Color Computer transmits 8 data bits, two stop bits, and no parity. This agrees with the LA50's factory setting, obtained by opening switches SW 2-4, SW 2-5, and SW 2-6. On the other hand, the LA100 requires odd or even parity if the printer is set up for 8 data bits. I got correct printing on the LA100 by configuring it for 7 data bits and space parity. This required setting the LA100's switches A6, A7, and A8 to the right. The LA100 I had could not print 8-bit characters, so using 7 data bits did not limit the printer further than using 8. Typical software on the Color Computer ends a line of text with a carriage return (ASCII decimal code 13) but no line feed (ASCII decimal code 10). A printer is expected to supply a line feed; otherwise, successive lines overwrite each other. I know of no way to make either an LA50 or an LA100 supply automatic line feeds. Instead, I configured the computer to send line feeds. Telewriter-128 has an option for this purpose. I have used the CoCo more often with the LA100 than with the LA50. The combination of inexpensive computer and fancy printer has produced nice-looking letters. ----------------------------------------------------------------------------------- How To Fix The XON/XOFF On Your LA50 Printer By K. H. Hawkins Modified slightly by Frank Mallory and David Maroun From the Boston Area Rainbow News, Summer 1988 Has your LA50 ever pooped out? If it hasn't, it probably will sometime. Mine did recently, after 3 years of use, but was still only on the second box of paper. When my word processor, SAMNA WORD II, displayed the error, "Something is wrong with your printer. Please check it", I performed a self test, but nothing showed amiss. When I tried the command, "PIP LST:=test_file_name", the printer printed for 56 lines or so, and then produced row after row of (backward) question marks. Service advisors who have service contracts for large industrial users say 80 - 90% of all LA50 problems trace back to a capacitor on the mother board. Replacing this capacitor has saved sending the printer off to be serviced at the main plant. Before you spend $100 - $150 on servicing the printer through DEC or a third-party company, risk a dollar and spend an hour trying this technique. (1) Get a 100 microfarad 25 volt electrolytic capacitor from your local electronics store. Radio Shack stocks these. It should be about 3/16" in diameter and 3/8" long. (2) Disconnect the LA50 power cord and printer (RS232) cable and remove any paper. To avoid damaging the clear plastic flap on top of the printer, remove the printer cover by grasping the top cover and sliding it slightly up and toward the front of the printer. Next, turn the printer upside down on a soft cushion, with the front of the printer facing you. (3) Remove the four phillips-head screws holding the bottom access plate, and remove the plate. Remove the four screws securing the printed circuit board and gently pry up on the left side. Disconnect enough cable connectors to allow you to swing the left side of the board up and get access to components in the center of the board. You need not worry about connector identification because they are all unique in size and position. (4) Locate two 100 microfarad 10 volt electrolytic (tubular) capacitors side by side near the center of the back edge. Unsolder the one closest to the printer (i.e., at location C) and solder the new one in place. Take care to observe that electrolytic capacitors have polarity--don't put it in backwards. Note that you are replacing a capacitor rated at 10 volts with one rated at 25 volts. This is so you will have to do this procedure only once! (5) Now repeat the above disassembly in reverse order. You will need a bit of patience aligning and pressing the connectors back in place. The whole procedure should take about one hour. ----------------------------------------------------------------------------------- Review: Mace Utilities, Version 5 By Wilson C.Y. Chang Peter Norton has a challenger these days. Paul Mace's disk utilities are often featured in advertizements as an alternative to Norton's. Both Norton and Mace utilities duplicate what is supplied with operating systems, or is available in the public domain or as shareware. However, in many an office I have seen, it is almost unthinkable to use cheap software when expensive stuff is around. After all (the theory goes), if you do not constantly spend large amounts of money, you may not get any when you really need it. So, I tried version 5 of the Mace utilities. I ran them on a Rainbow 100A equipped with 852 000 characters of random access memory and an NEC V20 microprocessor but no hard disk drive. The Mace manual was easy to read and thorough in its explanations. The authors anticipated many problems that might arise, and provided solutions in a question-and-answer format. In fact, the manual starts off with problem-solving. I suppose this format might appeal to people who are anxious to get a minimum of information needed to use the utilities. I recommend reading the whole manual at least once. A README file is supplied on disk with last-minute information. I also consider it required reading. The Mace utilities are not copy-protected. In fact, the manual stresses making copies of the master diskettes, and keeping the masters write-protected and safe. Very good advice, I say. The manual gives a telephone number to call in case there is a serious problem which a user cannot solve without help. I did not try calling. Using The Utilities Included with the Mace utilities is INSTALL.EXE, a program for installation on a hard disk. I did not have a hard disk, but I tried INSTALL anyway. It ran nicely under Rainbow MS-DOS 2.11-1 on the 100A. An experienced user might just use the MS-DOS 'COPY' command to put whatever utilities seemed most interesting onto a hard disk. MACE.EXE is another program provided more for convenience than necessity. MACE.EXE is a menu program for calling up some (not all) utilities, and corresponds to Norton's NI.EXE. MACE.EXE also provides information about the utilities. Menus may be helpful to a beginner, but an experienced user can get faster results by calling a utility from the MS-DOS prompt. MACE.EXE has options for using an IBM PC-type monochrome display or ANSI screen controls. When I tried the monochrome option, the program used only a 40-column screen width; the right half of the screen display was missing. I could not get the ANSI option to work at all. Best results came from using the IBM PC-DOS emulator Code Blue with its '/C' option and MACE.EXE in its default color mode. MACE.EXE ran slowly, a fault it shares with Norton's NI. FFIND.EXE will search an entire drive for a file, provided you have some idea what its name is. A '/D' option allows searching through deleted files. FFIND ran without any noticeable problems under Rainbow MS-DOS 2.11-1. However, FFIND has many competitors. There are plenty of "WHEREIS" programs in the public domain or shareware realm. Some of these competitors come with their source codes, some can find files that are in various kinds of archive, and any I know of are much smaller than FFIND's 33 491 characters. For example, the shareware SFIND.COM has a mere 2510 characters, can read directories of .ARC files, and is dazzling fast. To my knowledge, though, FFIND's '/D' option is unique. FORMAT-F.COM is listed as a safe formatter for floppy diskettes. The program erases the root directory and file allocation table, but restores all other information to a diskette after formatting it. Thus, an unformat program can restore files if need be. FORMAT-F.COM ran under Rainbow MS-DOS 2.11-1, but would not format in an RX50 drive. However, the FORMAT command included with Rainbow MS-DOS, if used without its '/I' option, also leaves information intact except the root directory and file allocation table. I successfully restored files on Rainbow diskettes that had been formatted without the '/I' option. FORMAT-H.COM formats hard disks just as FORMAT-F.COM does floppy diskettes, erasing only directories and file allocation tables. The program is important only to users of certain computers with certain versions of the MS-DOS FORMAT command. COMPAQ owners, for example, might appreciate FORMAT-H.COM. Most versions of the format program supplied with MS-DOS or IBM PC-DOS preserve the main parts of files on hard disks. I did not test FORMAT-H.COM on a hard disk. If a file is edited often, it is likely to become fragmented; that is, the file becomes broken into pieces located on various parts of a disk rather than in one place. Such fragmentation can slow down access to the file since disk heads must search to find the pieces. FRAGCHK.EXE searches an entire disk and tells what files are fragmented and how many fragments exist. However, the CHKDSK utility supplied with MS-DOS gives that information if used in the form CHKDSK B:*.*<Return> (if B: is the drive being checked for fragmented files). Unlike FRAGCHK, the 'CHKDSK *.*' command must be repeated for each subdirectory on a disk. You can decide for yourself whether that difficulty is enough to make you buy FRAGCHK. Anyway, FRAGCHK ran as intended under Rainbow MS-DOS 2.11-1, except that it used the first file on a disk as the volume label. VACCINE.EXE is supposed to protect the system area of a disk from modification, such as attacks by viruses or accidental disk formatting. When I tried to run VACCINE, I was told it was already installed. Thus the program proved ineffective, at least on the Rainbow I had. Disk caching is using of part of memory as temporary storage to avoid unnecessary disk accesses. Caching is supposed to speed operations, especially on floppy disks. Three disk caching systems are supplied with the Mace utilities, one for standard memory, one for Lotus-Intel-Microsoft expanded memory, and one for AT extended memory. Each system consists of a driver to be added through a CONFIG.SYS file. A program, CACHCTRL.EXE, can be used to evaluate the advantage of disk caching. I used the driver for standard memory. To get CACHCTRL to give its intended screen display, I needed Code Blue. The disk caching system seemed to work, but the speed gain was very small--about 0.1%, according to CACHCTRL. Disk caching is built into MS-DOS. By specifying 'BUFFERS=6' in a CONFIG.SYS file, you set aside part of memory (expect about 3 000 characters, but the amount varies from one system to another) for temporary storage. You can change the size of the cache by varying the number in the BUFFERS assignment. I ran CACHCTRL with 'BUFFERS=6' in CONFIG.SYS, and with 'BUFFERS=0' to disable caching. Again, the difference in speed was negligible. Since doing that experiment, I have changed the BUFFERS assignment in my CONFIG.SYS files to the smallest number available, 0 under MS-DOS 2.11-1 or 1 under MS-DOS 3.10 on Rainbows. Why waste memory? As you have probably guessed, I do not use the Mace disk caching systems. MKEYRATE.COM is intended to adjust the keyboard repetition rate and delay time for IBM ATs and PS/2s. This program had no noticeable effect on a Rainbow, and I did not get a chance to try it on an AT or PS/2. MUSE.EXE is a disk editor, a program for reading a disk sector by sector and making changes. The name 'MUSE' is an acronym for 'Mace Utilities Sector Editor'. In one respect, MUSE outdoes any similar program I have ever tried: MUSE is the biggest. MUSE occupies all of 201 128 characters on disk. The corresponding Norton utility, version 4, advanced edition, is NU.EXE, which has 124 641 characters, while its help file takes up another 10 897. Version 3.1 of NU has only 58 096, and the CP/M-80 DU.COM version 8.9 has a mere 7 680. MUSE offers various options for configuring the program to different computers. On the Rainbow 100A, I got the results I liked best by using Code Blue and MUSE's '/D' option. Specifying '/D' tells MUSE it is running on a machine which is compatible with an IBM PC in its basic input-output system, but not in hardware, and has a monochrome display. Response to my keystrokes was sluggish. Otherwise, I noticed no special problems with MUSE. It seemed to handle the Rainbow MS-DOS diskette format well enough. PARK.EXE is supposed to park hard disk heads, and was a subject for an editorial in the VARUG newsletter not long ago. The editorial mentioned that the program tried to park any drives labeled 'C:' or higher. I confirmed that PARK.EXE would try to park a RAM drive, which was E: on the Rainbow I used. This bit of nonsense is, I guess, good for a laugh. I did not try PARK on a hard disk. REMEDY.EXE will search all of a disk for bad sectors, move any information off the bad sectors, and lock out those sectors. REMEDY is a rather big program: 98 791 characters. I used REMEDY on the Rainbow with Code Blue and Code Blue's '/C' option for color displays. If you do not have REMEDY, you can lock out bad sectors by copying all files from a suspect disk and reformatting. If you use Rainbow MS-DOS 2.05, 2.11, or 2.11-1, you should not use the FORMAT command's '/I' option. If you have a suspect hard disk, the safest procedure is to back up all files and do a complete low-level format and partition. I know that is a time-consuming task with a large hard disk, but you probably will not have to do it often. The software needed for low-level formatting is typically provided with hard disks. Rainbow users who have the standard DEC hard disk controller can also use the public-domain WUTIL program. RXBAK.EXE saves a copy of a disk's boot sector, file allocation table, and root directory in a file called 'BACKUP.M_U'. RXBAK is supposed to be used with Mace's UNFORMAT.EXE to recover files in case a disk is accidentally formatted. The two programs are designed to work with either a hard disk or diskette. I did not care for RXBAK's storing BACKUP.M_U on the disk of interest. I would prefer to have this file on another disk besides the one I was trying to protect. Furthermore, the RXBAK-UNFORMAT combination shares a problem with other such systems: For best results, you must run RXBAK every time you change a directory on any disk you use. I cannot imagine anyone taking the trouble to do that! Anyway, RXBAK ran well enough under Rainbow MS-DOS 2.11-1 if COMMAND.COM was available. UNFORMAT ran best under Code Blue; otherwise, I got a garbled screen display, and sometimes I had to reset the computer after a run. Even with Code Blue, UNFORMAT sometimes would not return the computer to the MS-DOS prompt, but entering control-C set matters aright. SORTD.EXE rewrites directories in a specified order--in alphabetical order of filenames or extensions, for example. Either ascending or descending order can be used. If COMMAND.COM was available, SORTD.EXE ran without problems and without Code Blue. SQZD.EXE removes references to deleted entries from directories. The program required COMMAND.COM's presence, but worked as intended under Rainbow MS-DOS. SYSTAT.EXE is supposed to give technical information about its host computer. It ran without Code Blue, but indicated only 1 floppy drive and one serial port, and did not recognize memory beyond 640 k. SYSTAT correctly identified the microprocessor as a V20. UNDELETE.EXE recovers deleted files. This program required the presence of COMMAND.COM, and successfully recovered deleted files. However, those deleted files were unfragmented. The real test of any undelete command is to have it recover fragmented files, and I did not try that. The manual recommends MUSE for recovering fragmented files. If UNDELETE.EXE was given the name of a single file to recover, the program ran without Code Blue. If an ambiguous name (like '*.EXE') was supplied, then Code Blue with its '/C' option was needed to get a correct display and to see UNDELETE's prompts. UNFRAG.EXE unfragments files. This program can work with REMEDY.EXE to examine a disk first and then move clusters around until all pieces of each file are beside each other. By default, UNFRAG "packs"--puts files near the edge of the disk so that no unused clusters are between used ones. This speeds up disk access. To get a correct screen display from UNFRAG on the Rainbow, I used Code Blue with its '/C' option. UNFRAG took some nine minutes to defragment a floppy disk. This is a fairly long time compared to that taken by, say, version 2.06 of Disk Organizer (DOG), a shareware defragmenter. When UNFRAG finished, it reset the computer. The reset is just what the Mace manual says should happen, but I utterly fail to understand why it is done. I like to defragment a hard disk just before finishing work, then park heads and turn the computer off. If I used UNFRAG on a hard disk, I would have to load MS-DOS again just to park heads, and that would be a nuisance. You can always defragment by copying all files to other disks, erasing the original disk completely, and then copying the files back one by one (not by using DISKCOPY). Once a disk is defragmented this way, on later occasions you need deal only with newly-added files, and the process takes less time. If you ordinarily work on floppy disks or with small files on hard disks, you are not likely to notice any difference in the time taken to load a file when it is fragmented compared to when it is defragmented. The main advantage of defragmenting in these cases is to ease recovery of accidentally deleted files. Maintaining recent back-up copies can eliminate even that advantage. Conclusion The Mace utilities should be considered along with the alternatives, some of which I have pointed out in this article. There are still other alternatives; browsing through a catalog of shareware and public domain software can reveal them. ----------------------------------------------------------------------------------- A DEC-Oriented Bulletin Board Gary Stebbins' Glacier Peak bulletin board is the semi-official board for Washington State's d:BUG users group. Glacier Peak features files of special interest to users of Digital Equipment Corporation computers. Glacier Peak can be reached by dialing (206) 644-8431, and accepts calls at 300, 1200, and 2400 bits per second. Here is an edited partial list of files on the board. The list is dated December 13, 1989, but represents reasonably well what is on the board as of March 1, 1990. By now, many .ARC files have been converted to the ZIP format. ---------------------------------- Download Area #1--General Interest Files AIMODEM2.ARC 92672 09/04/87 An artificial intelligence communication program ASMBLUE.ARC 82511 09/25/86 Routines from Waite's BLUEBOOK of ASM language ASM_PRIM.ARC 32768 09/05/86 Source to assembly language primer by Lafore (archive is truncated--last file bad) DECBBS.LST 8170 11/27/89 List of Rainbow Fido BBSes F343-003.BBS 0 12/13/89 Summary of all files available on Glacier Peak Rainbow F343-003.ARC 7140 12/12/89 Archive of F343-003.BBS FIDO650.NWS 28120 12/11/89 Latest Fidonews FONT.ARC 14720 04/24/88 Terminal font for DEC windows INDX-343.ARC 213344 06/15/89 List of files on BBSes in Net 343 206LST06.89 26977 06/09/89 Western Washington BBSes June 1989 Download Area #2--d:BUG Related Files DBUG_ASC.ARC 113216 09/16/89 1988-9 d:BUG newsletters, plain text DBUG_WP.ARC 128096 09/16/89 1989-9 d:BUG newsletters, WordPerfect format Download Area #3--MS-DOS Files This area contains the general MS-DOS files. The programs located in this area should run on any MS-DOS (version 2 or later) machine. Most files are contained in archives (file type .ARC). ARC is needed to unpack and maintain these archives. The latest version of ARC is itself an archive containing the executable and documentation. Just execute it to unpack it. (You need about 100 k free space on your disk to unpack it.) I'm now converting many files to .ZIP format. Get PKZ102.EXE to unpack these files. When executed, PKZ102 will unpack itself into several files, including documentation. ADD_INS.ZIP 51995 02/21/89 1-2-3 version 2.01 add-ins for Rainbow ARC522.EXE 87040 08/01/88 Version 5.22 of the SEA arc|de-arc program ARCA129.ZIP 10132 06/26/88 Small ARC creator ARCE31C.ZIP 8683 09/16/87 Small ARC extractor ASK3.ZIP 17668 12/11/86 Batch file question asker ASMGEN22.ZIP 17582 06/11/86 Good disassembler--version 2.02 DCOPY.ZIP 13493 04/07/86 Disk COPY--allows disk image to be copied DIRTYDOZ.ZIP 22441 05/15/88 List of Trojan programs DOG206.ZIP 35488 03/06/89 Disk OrGanizer version 2.06 INTRCPT.ZIP 5477 04/13/87 Allows using 2S2D diskettes in 1.2 MB disk drive KER_SCP1.ZIP 14527 09/11/88 Sample Kermit script files KMICRO.ZIP 10823 04/09/85 Kermit usage article LU-SQ.ZIP 32539 04/09/85 Library utility (needed for *.LBR files) and SQ/USQ (squeeze/unsqueeze) utilities LHARC113.EXE 34305 05/21/89 Another archive program (.LZH files) MSKERDOC.ZIP 157896 02/15/89 Documentation for version 2.32 MS-DOS KERMIT MSVIBM.ZIP 67168 01/24/89 Kermit version 2.32a for IBM PAK210.EXE 91392 09/12/89 ANOTHER archive utility (.PAK files) PK361.EXE 119598 08/04/88 Version 3.61 of PK's PAK utilities (.ARC files) PKZ102.EXE 136192 10/16/89 PK ZIP version 1.02 (.ZIP files) RBE.ZIP 26295 02/01/87 Rainbow emulator for IBM machines RDIR.ZIP 30543 01/09/86 Extended directory program RX50.ZIP 18835 09/10/86 Allows IBM AT to read/write DEC RX50 diskettes SEZ230.EXE 9902 08/07/87 ZOO self-extracting archive builder SHAR.ZIP 25376 07/22/86 Unix-like Shar/Unshar utility SHOWMEM.ZIP 10653 07/30/88 Version 1.2 w/bug fixes & /Verbose option TOUCH.ZIP 7620 09/27/87 Set file date/time XCHG_104.ZIP 13759 09/13/89 Convert Archive formats XON2.ZIP 3376 12/19/88 Allow XON/XOFF operation on PC-compatibles ZOO201.EXE 82486 08/25/88 ZOO archive utility (.ZOO files) MAILCK26.ZIP 52247 08/30/87 OPUS Mail checker Download Area #5--Turbo Pascal Files INLINE.ARC 34124 02/11/86 Compiler to produce inline Turbo code FASTVID.ARC 3106 04/09/85 Fast video routines for Turbo KINSTV2.ARC 17494 12/16/86 Key install for Turbo on Rainbow TURBOKEY.ARC 7614 08/20/85 Key reader for Rainbow Download Area #6--C Files This area contains programs and routines written in C. CB.ZIP 9250 08/06/86 C beautifier ENDECODE.ZIP 14768 03/31/86 Encode and decode utilities--compatible with Unix UUENCODE and UUDECODE. Allow binary files to be transmitted over non-binary communications lines. PROBE.ZIP 36088 08/10/86 DOS version of UNIX's profile WINDOSYS.ZIP 35360 03/25/86 Windows for DEC Rainbow (sources) AREACD30.ZIP 18499 10/11/86 Area Code finder (with C source) Download Area #7--Rainbow games CHESS.ZIP 6904 05/04/86 Nice--uses VT-100 graphics GOTELO.ZIP 29667 05/17/86 Fast & smart Othello. Great graphics w/ card. QIX.ZIP 17282 04/29/87 Like the QIX Arcade game WORM.ZIP 15283 04/12/87 "Worm" game YAHTZEE.ZIP 21371 10/19/86 Dice game for RB--uses VT-100 graphics SOLIT.ZIP 34836 06/08/89 Game of Solitaire, needs graphics, nice! Download Area #8--SEDT (Screen EDiTor) Screen EDiTor version 4.0(260). Excellent editor. Fully tailorable. Has some word processing features, plus programming features. Executables here are for Rainbows & IBM compatibles, VAX/VMS, and ULTRIX (VAX version and DECstation 3100 version). Please note: SEDT v. 4.0 is Shareware. Anker is asking for contributions so he can afford to continue development. If you use SEDT, consider paying a little for it. SEDT4DOC.ZIP 75835 10/23/89 Documentation Files SEDT4DOS.ZIP 74458 10/23/89 MS-DOS executable SEDT4KEY.ZIP 62760 10/21/89 MS-DOS key and help Files SEDT4VMS.ZIP 238809 10/27/89 VAX/VMS Version SEDT4ULT.ZIP 26285 10/26/89 Ultrix key and help Files SEDT4VAX.ZIP 179479 10/26/89 Ultrix executables for VAX SEDT4MAX.ZIP 183258 10/26/89 Ultrix executables for DECstation 3100 SEDT4WPS.ZIP 47681 09/26/89 WPS key mapping files (possible problems) SEDT3DOC.ZIP 75460 02/07/89 SEDT v. 3.3 documentation SEDT3DOS.ZIP 71476 02/08/89 SEDT v. 3.3 for MS-DOS (Rainbow and IBM) SEDT3KEY.ZIP 66435 12/05/88 SEDT v. 3.3 key files for MS-DOS Download Area #11--DEC Rainbow 100 Files ACV.ZIP 17539 07/14/86 Translate characters: DEC to IBM and back. V. 1.2. ALARM.ZIP 13724 07/19/88 Alarm clock--includes sources. AME86-72.ZIP 57833 06/08/88 Run CPM-86 programs under MS-DOS. Version 7.2 w/source. AMORT010.ZIP 16893 09/13/88 LOTUS 1-2-3 spreadsheet for amortization tables AUTOPARK.ZIP 3803 08/27/87 Parks hard disk heads after inactivity--version 1.1. BOOT.ZIP 22104 02/05/85 Boot another hard disk partition. BYE.ZIP 526 11/03/85 Park hard disk heads. Everyone should use this. CABLE.TXT 6036 12/15/86 How to build dual monitor cable for Rainbow CAL.COM 1024 10/25/88 Simple calendar: CAL MONTH YEAR (default: This month) COUNTRY.ZIP 14644 02/22/85 Show country information about your system. DECWPRIN.ZIP 8904 04/30/87 WordPerfect LA50/100, LN03 printer drivers DIBEMV3.ZIP 13162 06/07/86 Version 0.3 of Dan Pleasant's IBM emulator DRAW14.ZIP 53776 12/29/86 Good graphics drawing program for Rainbow DSKPREP.ZIP 80159 08/05/86 Disk preparation tool--writes secondary boot. ENVIRON.ZIP 1149 03/01/86 Increases environment to 2 k. Works! FANCYEXE.ZIP 40502 06/18/85 FancyFont program--executables FANCYFON.ZIP 37145 12/29/85 FancyFont program--fonts FORIBM.CPM 3840 05/02/86 Formats IBM diskettes on the Rainbow FV.ZIP 5094 02/26/85 FAST-VUE--fast-video viewing of text files GW_DUMP.ZIP 3683 05/10/87 Dump graphics screens. HARDDISK.ZIP 7762 02/23/88 Patch Rainbow hard disk error handling HAYES.SCR 2279 02/10/88 LCTERM script to dial Hayes modems HISTRY54.ZIP 10614 03/27/88 Bryan Higgins' HISTORY version 5.4 for the Rainbow JOBSV41B.ZIP 40809 04/10/88 Version 4.1 with bug fix of JOBSDUMP for Rainbow LA50.ZIP 7863 03/07/88 Printer driver for LA50/75 under MS WINDOWS LA50FIX.ZIP 1776 02/15/89 How to fix LA50 printer LCTERM.ZIP 47326 04/28/85 Terminal emulator and file transfer program LDCOPY.ZIP 6909 01/10/86 New loaders--included with MS-DOS v. 2.11-1, fixes some bugs. If you have MS-DOS v. 2.05 or v. 2.11, then get this and fix your diskette access problems. MAINT.ZIP 23213 01/27/86 Directory display and maintenance utility MSVRB1.ZIP 51728 01/24/89 Kermit V2.32a for Rainbow. Docs in MS-DOS area. MSVRB1B.ZIP 51239 10/08/89 Fixes F11-F13 mapping and modem signal control. MSVRB1C.ARC 65153 11/11/89 Fixes printer control (updated 11-Nov). MSVRB1C.ZIP 53428 11/11/89 Fixes printer control (updated 11-Nov). MULTIPLN.ZIP 22137 11/13/87 MULTIPLAN version 3, 132 and 80 column for Rainbow OPTION.ZIP 6317 12/06/86 Display Rainbow system options. PAT31RB.TXT 2830 11/20/87 Patches to Rainbow MS-DOS 3.10. PRTSCR.ZIP 4192 07/12/85 Print the text on your screen at any time. RB-BUFFR.ZIP 17325 05/24/87 V. 3.0; expands Rainbow type-ahead buffer to 128 characters, and improves ^C handling. Includes AUTOTYPE. RBCONVRT.ZIP 6323 04/23/88 Documentation on converting A 'bows to Bs RDCPM.ZIP 5194 01/04/85 Read CP/M diskettes RDIRV12.ZIP MISSING 1/04/85 Directory program, allows sorting of listing by date, name, type. RNP10D.ZIP 5104 07/31/86 Memory-resident notepad/editor/screen editor RPM.ZIP 4070 02/18/85 CP/M utility to check RX50 drive speed RR20.ZIP 1284 09/01/84 Reset Rainbow SECBOT31.ZIP 13009 06/19/88 Update to RB BOOT for ST-251 & MS-DOS 3.10 SKI12A.ZIP 22028 03/29/86 Sidekick imitator version 1.2A by Bob Brodt TC.ZIP 22418 01/20/86 Terminal emulator--VT220 functions VDEC_021.ZIP 4434 11/13/88 Rainbow video fossil driver WP80132 1678 02/10/88 BAT files: Use WordPerfect in 80/132 column mode WSS322RA.ZIP 90005 11/01/87 Diskette catalog program WUTIL32.ZIP 97032 08/15/89 Winchester disk utility version 3.2 WUTIL32S.ZIP 93138 08/15/89 Sources for above BI2.ZIP 40004 06/28/89 Improved version of BICENTENAIRE PIBTERM.ZIP 175170 01/22/87 Communications program for Rainbow RBGIF.ZIP 36659 08/27/89 Rainbow-specific GIF viewer from CompuServe WILEY.GIF 7680 10/18/88 Demo file for RBGIF TC2PATRB.ZIP 6144 10/07/89 Patch Kit for Turbo-C version 2 from G. Theall VR241.TXT 4692 10/15/89 VR241 cable for CGA use ----------------------------------------------------------------------------------- What Is Available: Seagate Hard Disks (This section provides information on products likely to interest users of personal computers. Information presented here is based on that given by suppliers, with some editorial comments.) Seagate Technology Incorporated has provided hard disks for many Digital Equipment Corporation computers. For example, DEC's ten-megabyte RD51 disk is Seagate's ST412 (which Seagate no longer produces), and the 20-megabyte RD31 is the ST225 (which is still available). The ST412's modified frequency modulation (MFM) recording method and interface remain industry standards used by many current disks, whether from Seagate or from other manufacturers. Also available now are run-length limited (RLL), AT, and small computer systems interface (SCSI--often pronounced "scuzzy") disks. More than one of these features may be on a given disk. For example, a disk recorded with the ST412 method may have MFM or RLL recording density. Present drives may be half-height (the ST225 is an example), or full-height (like the ST412). The diameter may be 5.25 inches (as for the ST225 or ST412) or 3.5 inches. Unformatted capacities range up from about 25 million characters (on the ST225). Drive specifications are indicated by the name. The following diagram tells how to read many Seagate model designations: ---------------------------------- ST251N-1 | \ \ \ | \ \ > Drive Speed: 1 corresponds to 40 milliseconds average access time | \ \ 0 corresponds to 28 milliseconds average access time | \ \ | \ >----------Recording Method: None corresponds to ST412/MFM | \ R corresponds to ST412/RLL | \ N corresponds to SCSI | \ A corresponds to AT | > Unformatted Capacity | (up to 3 digits) |-> 1 corresponds to 3 .5-inch, half-height -> 2 corresponds to 5.25-inch, half-height -> 4 corresponds to 5.25-inch, full-height ---------------------------------- The following table lists some Seagate hard disks with their interfaces and approximate formatted capacities in millions of characters: ------------------------------------------------------------ | | Approximate Capacity | | Interface | 20 30 40 60 80 | | ~~~~~~~~~ | ~~~~~ ~~~~~ ~~~~~ ~~~~~ ~~~~~ | | 412/MFM | 225 138 251 4096 | | | 125 | | | | | 412/RLL | 238R 250R 277R | | | 138R 157R | | | | | SCSI | 225N 138N 251N 277N 296N | | | 125N 157N | ------------------------------------------------------------ Different systems use different numbers of sectors per track, so actual formatted capacities may vary somewhat. Here are more detailed technical specifications for some Seagate drives. 'MTBF' stands for the mean time between failures. ST138N 1) UNFORMATTED_______________________________ - 2) FORMATTED (26 SECTORS)____________________32.2 MB 3) ACTUATOR TYPE_____________________________STEPPER 4) TRACKS____________________________________2 460 5) CYLINDERS_________________________________615 6) HEADS DATA/SERVO__________________________4/0 7) DISCS/TYPE________________________________2/THIN FILM 8) RECORDING METHOD__________________________RLL 9) TRANSFER RATE Mbits/s_____________________7.5 10) INTERFACE________________________________SCSI/RLL 11) TPI (TRACKS PER INCH)____________________824 12) BPI (BITS PER INCH)______________________22 430 13) AVERAGE ACCESS - ms______________________40/28* 14) SINGLE TRACK SEEK - ms___________________8 15) MTBF (HOURS)_____________________________70 000 16) POWER (WATTS)____________________________9 17) LANDING ZONE_____________________________AUTO PARK * ST138N-0/ST138-1 ST225N 1) UNFORMATTED_____________________________ - 2) FORMATTED (17 SECTORS)__________________21.3 3) ACTUATOR TYPE___________________________STEPPER 4) TRACKS__________________________________2 460 5) CYLINDERS_______________________________615 6) HEADS DATA/SERVO________________________4/0 7) DISCS/TYPE______________________________2/OXIDE 8) RECORDING METHOD________________________MFM 9) TRANSFER RATE Mbits/s___________________5.0 10) INTERFACE______________________________SCSI/MFM 11) TPI (TRACKS PER INCH)__________________588 12) BPI (BITS PER INCH)____________________9 827 13) AVERAGE ACCESS - ms____________________65 14) SINGLE TRACK SEEK - ms_________________20 15) MTBF (HOURS)___________________________100 000 16) POWER (WATTS)__________________________16.8 17) LANDING ZONE___________________________670 ST251N 1) UNFORMATTED_______________________________ - 2) FORMATTED (26 SECTORS)____________________43.1 3) ACTUATOR TYPE_____________________________STEPPER 4) TRACKS____________________________________3 280 5) CYLINDERS_________________________________820 6) HEADS DATA/SERVO__________________________4/0 7) DISCS/TYPE________________________________2/THIN FILM 8) RECORDING METHOD__________________________RLL 9) TRANSFER RATE Mbits/s_____________________7.5 10) INTERFACE________________________________SCSI/RLL 11) TPI (TRACKS PER INCH)____________________777 12) BPI (BITS PER INCH)______________________14 902 13) AVERAGE ACCESS - ms______________________40/28* 14) SINGLE TRACK SEEK - ms___________________8 15) MTBF (HOURS)_____________________________70 000 16) POWER (WATTS)____________________________12 17) LANDING ZONE_____________________________AUTO PARK * ST251N-0/ST251N-1 More information is available from Seagate dealers, or from Seagate Technology Incorporated 920 Disc Drive Scotts Valley, California 95066-4544 United States Of America Telephone (408) 438-6550 For sales, telephone toll-free 1-800-468-DISC (that is, 1-800-468-3472) FAX: (408) 429-6356 ----------------------------------------------------------------------------------- Questions And Answers Do you have a computer-related problem? Send it to us. We can publish it, and if we do not know a solution, perhaps someone else in the users group can provide one. QUESTION: I have two computers side by side that cannot communicate on diskette. One machine accepts only 3.5-inch diskettes, and the other only 5.25-inch diskettes. Both computers do have RS232C serial ports. Can you tell me how to use a modem to transfer files from one computer to the other? ANSWER: A modem is designed for communication over telephone lines, not for your situation. If computers are near each other, all you need is a cable and communications software. The cable should connect serial ports, and should include a null modem adapter to ensure that outgoing signals from one port reach the receiving pin on the other. You must examine your particular ports to know what connectors you need. Cables and connectors should be available from electronics or computer stores generally, though you may have to put them together yourself. You do not need the same communications software on the two computers, but they should use compatible protocols; that is, if you have YMODEM on one machine, you need YMODEM on the other. QUESTION: How can I delete a PC-FILE + report? The manual does not tell how. ANSWER: PC-FILE + saves its reports in files with the extension '.REP'. You can go to the MS-DOS prompt and delete any such file. To delete a report while running PC-FILE +, go to the main menu and choose utilities (press the <M> or <F8> key). When the utilities menu comes up, choose maintenance (press the <M> key). Say you want to delete (press <D>), press <R> for reports, select the drive that the report is on, select the path, choose the report, press <F10>, and type <Y> to confirm your choice. PC-FILE + will provide prompts for all steps. Directions for deleting reports are not in the manual for version 1.0 of PC-FILE +, though the program allows the procedure. The manual for version 2.0 does give directions. QUESTION: I often use a Rainbow in terminal mode. How can I record a communications session or a message? ANSWER: To record the entire conversation, turn on your printer and type <Ctrl> <Print Screen>. The printer will print out what is on the screen as the session goes on. To turn off printing, type <Ctrl> <Print Screen>; that is, the <Ctrl> <Print Screen> sequence acts as a toggle, like <Ctrl> <P> under CP/M or MS-DOS. To print only what is on the screen at the moment, press <Print Screen>. We have used this procedure on a Rainbow 100B with a printer that used XON-XOFF signals and did not provide an automatic line feed when it received a carriage return. The procedure was also successful with a printer which used the data terminal ready signal instead of XON-XOFF, and which did provide an automatic line feed when it received a carriage return. The second printer did not double-space. This result surprised us, since under CP/M or MS-DOS, a Rainbow requires different configurations for the two printers, and the second printer has to be told to turn off the automatic line feed to avoid double-spacing. A Rainbow 100A in terminal mode worked like the 100B with the first type of printer, but with the second printer, pressing <Print Screen> printed out only one line. We also got double-spacing with that second printer. QUESTION: Sometimes when I use the GW-BASIC interpreter, I would like to record the session to a file. How can I do that? ANSWER: When you load the interpreter, pick a file name (say, 'RECORD.DOC'), and type GWBASIC > RECORD.DOC<Return>. You will record the session to RECORD.DOC, but will still get a screen display. We have used the same procedure successfully with programs compiled by the GW-BASIC compiler. QUESTION: How do I dial out from a Hayes-compatible modem? I have a dialing program which dials out correctly, and then calls up the communications program I want to use, but that procedure limits what I can do with the communications program. ANSWER: You can just load your communications program, go into terminal mode, turn on the modem, and type in a Hayes dialing command. If the number you want to dial is 100-0000, and your telephone uses touch tones, type ATDT 100-0000<Return>. The space before the '1' is optional, as is the '-'. You are safest to use only upper-case letters for 'ATDT', but many modems will accept lower-case too--but not a mixture of upper- and lower-case. The command 'Atdt', for example, will likely be rejected. If your telephone uses pulses instead of tones, type ATDP 100-0000<Return>. You may find that you can use simply 'ATD' rather than 'ATDT' or 'ATDP'. If your modem does not have an external volume control, we suggest you precede dialing with the command to turn down speaker volume: ATL1<Return>. ----------------------------------------------------------------------------------- Buy, Sell, Or Swap This section is presented as a service to members. There is no charge for advertizements placed here, though donations will be accepted. Only items related to computing will be advertized; if you wish to sell an old car, we respectfully suggest that you publicize elsewhere. Advertizements are not accepted from suppliers. In accordance with DECUS policy on commercialism, we do not print prices. Ads should preferably be submitted to the editor in writing or as ASCII computer files, but may also be phoned in. ---------------------------------- FOR SALE: VAX 11/730 unlimited user with MicroVAX II as end node Rainbow 100Bs with keyboards, monochrome monitors, 10 megabyte hard disks, and 256 k or more of RAM Rainbow 100Bs with keyboards, color monitors, 10 megabyte hard disks, and 256 k or more of RAM DEC Rainbow version of dBASE III version 1.0, unopened Contact Charles Haynes at (604) 985-6125 during business hours. FOR SALE: Poly-XFR CP/M communications software for Rainbow 100; CP/M-86/80 operating system version 2.0. Both software items are in the original packages and have documentation included. One AC fan for a Rainbow 100A (also fits many other computers). 7.6 cm (3 inch) adding machine rolls. These items are just taking up space now, so all offers will be considered. Telephone David P. Maroun at (604) 792-4071. FOR SALE: One RX50 drive in good condition. Also, three 65 536-character DRAM (memory) chips. Telephone David P. Maroun at (604) 792-4071. WANTED: The manual for version 1.5 of the WPS-80 word processor. Any information about the communications sub-menu would be welcome. Contact Doug Nicol at telephone number (604) 792-0025. FOR SALE: Peachtree business modules for MS-DOS: PeachCalc spreadsheet, general ledger, accounts payable, accounts receivable, personal calendar, job cost system, and inventory control. Any offer will be seriously considered. Note: These modules require Code Blue and maximum memory to run on Rainbows. Contact David P. Maroun at (604) 792-4071. FOR SALE: DEC Rainbow 100-A1 system unit with an 11-megabyte hard disk, CP/M-86/80 version 2.0, MS-DOS version 2.05, and hard disk management utilities. Contact David P. Maroun at (604) 792-4071. ----------------------------------------------------------------------------------- What Do You Think Of This Issue? Please tell us what you liked and did not like. The best articles were:__________________________________________________________ _________________________________________________________________________________ _________________________________________________________________________________ _________________________________________________________________________________ The worst articles were:_________________________________________________________ _________________________________________________________________________________ _________________________________________________________________________________ _________________________________________________________________________________ Comments or suggestions: Send your opinions to The Editor, VARUG Newsletter, 9395 Windsor Street, Chilliwack, BC, Canada V2P 6C5.