Info-IBMPC@WSMR-SIMTEL20.ARMY.MIL (05/13/89)
Info-IBMPC Digest Sat, 13 May 89 Volume 89 : Issue 52 Today's Editor: Gregory Hicks - Chinhae Korea <COMFLEACT@Taegu-EMH1.army.mil> Today's Topics: New OS/2 Discussion Forum Sending binary across Bitnet using Ascify Public domain FF - Fast File Finder Program Need assistance with PC/IX Turbo Pascal to Turbo C Link Floppy Status Help - Print Position Aid Needed How to add a Hard Disk to an IBM PC Portable Illegal opcodes Re: BYTE Magazine Problems Laser XT/3 Mouse book vs. manual Passing parameters to batch files PK's ZIP program printer problems Toshiba 1.44MB Floppy Question TSR for Periodic execution of a Function (2 msgs) TSR proc. Turbo Pascal 5.0 Bug reports. Windows 386 problem ---------------------------------------------------------------------- Date: Tue, 09 May 89 10:13:03 MET From: Karel Van der Haegen <TEWKVDH@BLEKUL14> Subject: New OS/2 Discussion Forum OS-2@BLEKUL11 OS/2 Discussion Forum The purpose of the hereafter presented discussion forum is to gain information on, extend the knowledge about, and increase the insight in the fundamentals of The IBM/Microsoft OS/2 Operating System for Personal Computers more specific The OS/2 Standard Edition 1.0 The OS/2 Standard Edition 1.1 The OS/2 Extended Edition 1.0 The OS/2 Extended Edition 1.1 with The Presentation Manager (available in OS/2 1.1, the windowing concept) The Communications Manager (SDLC, DFT, Token-Ring, Asynchronous, 3270 emulation, X.25,...) The Database Manager (interactive SQL query and report writing) The OS/2 1.1 Toolkit (application development including presentation manager applications) The OS/2 Technical Reference Bindings The OS/2 Lan Server (functionality, interface with external resources from PCLP domains, support for PCLP PC-Dos Requesters,...) To avoid noise, to keep this list interesting, OS-2 will be moderated by editors, with messages from participants being blocked in newsletters. Seperate parts for answers on previous questions and new unanswered questions will be achieved through editing. The only language supported will be English. The host computer (BLEKUL11) is part of the computing facilities of the K.U.Leuven University of Leuven, Belgium. The discussion is moderated by 2 OS/2 Lan Server Local Area Network Administrators. Dirk Rober belongs to the Department of Economics and the CES Institute (Centrum voor Economische Studieen, Van Evenstraat 2B, B-3000 Leuven) Karel Van der Haegen belongs to the Department of Applied Economics and the LIRIS Institute (Leuven Institute for Research on Information Systems, Dekenstraat 2, B-3000 Leuven) The fact that this discussion forum is moderated doesn't influence the procedures for subscribing to the mailing list, or submitting mail to the mailing list. The periodicality of the newsletter will be dependent of the amount of traffic to the os-2 mailing list. Newsletters however will not exceed 1000 lines of text. Newsletters will be numbered YYMMXX with a sequence number XX reset at the beginning of each new month. The archives of the OS/2 Discussion Forum will be available to all on LISTSERV@BLEKUL11 on a monthly basis from 8905 on. Possibly in the future public domain OS/2 code can be distributed by a Listserv machine as well. ------------------------------ Date: Mon, 8 May 89 08:50:12 PDT From: madler@Hamlet.Bitnet (Mark Adler) Subject: Sending binary across Bitnet using Ascify I had this problem sending binary stuff across Bitnet a few years ago, and at that time I wrote a program called ASCIFY that I have been using successfully since. It is available on SIMTEL20 (with source and documentation) under <MSDOS.FILUTL>ASCIFY13.ARC. What Ascify does is to convert binary files to a text format consisting of only printable characters in a set of 85 that have unambiguous EBCDIC translations. They are grouped in lines of 72 characters to further appease picky channels. Spaces are not included in the set of 85. The conversion uses one of two modes, a text mode with nearly 100% efficiency and a binary mode that converts four bytes to five characters, giving an efficiency of 80%. The mode is decided automatically for each 512 byte block, and for each block, a two byte CRC is inserted for error checking. Ascify will convert one or several MS-DOS files into a single ascified file, preserving the files' names and date/time stamps. The same program does the inverse conversion, restoring the files to their exact condition, or reporting if the data was corrupted beyond the capacity of the program to interpret it. (Ascify will ignore extra control characters or spaces that might have crept in during its travels over nets, so long as the CRC's for the files check out.) Ascify is in assembler, and so it is very fast. I typically will use PKPAK or PKZIP to compress the files first, and then use Ascify to "expand" them to a printable format. The conpression almost always beats the expansion, so that the resulting file is actuall smaller than the files being sent. Mark Adler bitnet: madler@hamlet arpa: madler@hamlet.caltech.edu ------------------------------ Date: Mon, 8 May 89 13:43:54 MET From: <RCSTLN@HEITUE5> Subject: Public domain FF - Fast File Finder Program Hoi, I'am looking for the newest release from FF from Keith P. Graham. My version is the one that is printed below. FF.COM fast file finder - with options. Version 2.08 copyright 1988 by Keith P. Graham Total elapsed time: 0:0:4.78 FF.COM finished. For the latest release of FF call PC Rockland BBS - (914) 353 2176 I've read the newest version is 2.09. Can someone mail it to me. A newr version is also alright. I can only use it in the uuencoded way. I need I because I maintain a bulleting board and it's one of the best file finders I know. Anyone knows how to buy PCED, ( Where to order for it) one of the best command line editors there is. It isn't possible to buy it here in Europe because no-one knows about it. >>Thanks Leo v. Nieuwenborg You can reach me by mail RCSTLN@HEITUE5 ------------------------------ Date: Tue, 09 May 89 23:07:43 EDT From: "Dr. A.Bykat" <BYKAT@UTCVM> Subject: Need assistance with PC/IX I have inherited a 1984 version of the PC/IX (Unix system III) - manuals and all. The system can be installed on the old IBM PC/XT (8088 chip). I have an AT&T 6300 (V20 - 8086 chip with a Seagate 251 40 Mb hard drive) and would like to install the PC/IX on it. Unfortunately, although the system runs on a pure blue XT, it does come up on the AT&T machine. In fact the standalone Shell takes over displaying its prompt ($$) but when I enter the install command the shell hangs rather than displaying the disk preparation menu. Does anybody know how to resolve this problem? Help will be really appreciated. Alex Bykat Acknowledge-To: <BYKAT@UTCVM> ------------------------------ Date: 7 May 1989, 11:07:09 EDT From: AL148859 at TECMTYVM Subject: Turbo Pascal to Turbo C Link Hello, I'm trying to write a syntax checking program in Turbo C 2.0, but the lexical scanner is written in Turbo Pascal 5.0. The lexical scanner is a procedure that returns the next token to analyze. Now, my question is: can I call the Pascal lexical procedure from Turbo C? And can I use global variables in both the Pascal lexical scanner and the Turbo C syntax procedure? Thanks for the help! Juan Gabriel Ruiz Pinto Electronic System Eng. AL148859@TECMTYVM ------------------------------ Date: Tue, 09 May 89 15:47:20 +0200 From: Andr'e PIRARD <A-PIRARD%BLIULG11.BITNET@CUNYVM.CUNY.EDU> Subject: Floppy Status > Does anyone have any idea how to check the status of a floppy drive. >For example, PC Tools Backup will be able to see when the door is opened >(360K, XT) and then when it is closed when prompting for a disk. I'd like >to be able to figure this out for another project I have in mind. Any >ideas out there ? You get the illusion PCTB can sense the door state. In fact, all it does is continuous reading and telling "not ready" status from "sector not found" (or something like that). The only other technique I could find to avoid spending much time at read attempts is to monitor the write protect status soon often to catch the on/off transitions when a disk is slipped in or out. You will find my code at SIMTEL in my diskettes duplicator program: <MSDOS.DSKUTL>POLYCOPY.ARC Andr). ------------------------------ Date: 9 May 89 15:46:00 EDT From: "CHIEF LARRY KELLY" <kellyl@gw2.hanscom.af.mil> Subject: Help - Print Position Aid Needed Can anyone help me identify some sources for a print position aid. Its a clear overlay with grids for different pitch, etc., that you lay over a pre-printed form to format printer output. I've seen them in the past, but of course when I saw it I didn't need it and now when I need it - I can't find it. Thanks in advance. ------------------------------ Date: Tue, 9 May 89 11:21:46 +0200 From: mbe@dde.dk Subject: How to add a Hard Disk to an IBM PC Portable In comp.sys.ibm.pc.digest you write: >Anyone know how to add a hard disk to an IBM Portable PC? I have done this on 2 PPC's (original IBM's). I used the Hardcards from Plus Developments (the compagny that started the 'Hardcard' idea). The nice thing with this drive is : - It only takes one (1) slot in the PPC (full-lenght !) - Low power consumtion - Fast access (ca. 45 msek average seek time) - It makes a small blinking sign on the screen when accessing the drive. - Rather high g-handling (30 g's if I remember rigth) and autoparking. - It does not generate much acustic noise as most drives do. You probably has to change your display card to a half-length card to make space, but these (standard CGA cards) are not very expensive. The drive I used was the 20 MB type. Plus has also made an 40 MB drive, but I don't know if it can fit in the PPC. The drive was placed as the first card after the power supply, after that came a multifunktion full-length card and the 'nearly full-lenght' slot was used for the floppy controller card. The 2 hardcards have been used for 3 years with no problems at all. Hope this helps. Kind regards Martin Berg mbe@dde.dk | "The answer is 42" or | A. Douglas ..mcvax!enea!dkuug!dde!mbe | mbe@dde.dk or ..mcvax!enea!dkuug!dde!mb Disclaimer: As always I do not speak for my employer. ------------------------------ Date: Tue, 09 May 89 16:04:49 +0200 From: Andr'e PIRARD <A-PIRARD%BLIULG11.BITNET@CUNYVM.CUNY.EDU> Subject: Illegal opcodes The 80186, like the V20 has an instruction set half way between the 8086 (or 88) and the 80286 (but the V20 does not catch illegal opcodes). Two reasons I can see why they might occur: 1) The program tests the processor type (e.g. with a PUSHA (286 & 186)) and goes on with other 286 but non-186 instructions. You have to patch out the test. 2) The program is buggy with illegal opcodes, but they are harmless to a 8086. You have to debug. That's not counting the occasions where you see a message for an illegal instruction trap when others only observe a blind crash. Then, you must feel happy. But I think these events should be rare. Most programs keep to the 8086 instruction set. Is your processor healthy? Andr). ------------------------------ Date: 8 May 89 12:22:00 PDT (Monday) From: "George_C._Burkitt.ElSegundo"@Xerox.COM Subject: Re: BYTE Magazine Problems ><PAAI%HTIKUB5.BITNET@CUNYVM.CUNY.EDU> >I'm angry at BYTE. Very angry. >" subscribing and processing a change of address..." Well, it only takes 60 - 90 days for US subscribers to get entered, cancelled. I think the whole subscription process is handled by a separate operation from the magazine proper...it often is in a separate city, etc. Maybe a contract operation? At any rate, you are not alone. >"they ask for payment of the next subscription half a year before..." That's not a demand for payment, it's a request for renewal. Your new subscription period will be sequential with your old subscription. If you wait till the end of the present subs. to renew, you may invoke the 3 month delay again. This is a case of marketing folks not wanting to lose a customer. They do run the risk of getting a renewal every time they request (which turns out to be twice each year for a yearly subs. It is very common for a reader to let his subs. expire and then have to renew again as soon as the he finds out he is not getting the mag. This means two subs. actions for the mag distributor. They try to prevent this by reminding you twice; once early, once late. 'Tain't dishonest; they think they are doing you a favor as well as cutting their costs. Usually you can decipher your subs. ending date by reading the address label; otherwise you need to keep track of when you ordered what. Whichever renewal request you respond to (first, last or both) you will probably get all the issues you pay for. You or your heirs. ------------------------------ Date: Mon, 8 May 89 15:05 N From: <BRINKGRE%HLSDNL5.BITNET@CUNYVM.CUNY.EDU> Subject: Laser XT/3 I have a Laser XT/3 computer. I changed the origional 8086 uP for a NEC V30 processor. A few programs however refuse to work with the V30, so I would like to build something as a ' processor' switch to choose between the 8086 and the V30. Does anybody know of a reliable way to do this ? After a reboot Norton's System Info tells me that I have a 64K EGA. After I run a testprogram that came with the card SI tells me it is a 256K EGA. The card is a 256K EGA 4+ card from VideoTechnology. Does anybody with the same computer have this problem ? Or a solution ? If you don't have this problem, please mail me as well. Then i'll visit my dealer. Thanks in advance. Michel Brinkgreve Dr. Neher Laboratories, Netherlands PTT. Dr. Neher Laboratories, BITNET/EARN: BRINKGREVE@HLSDNL5 PO Box 421, SURFnet: DNLTS::BRINKGREVE 2260 AK Leidschendam, Netherlands. ------------------------------ Date: Mon, 08 May 89 17:44:04 PDT From: JAJZ801%CALSTATE.BITNET@CUNYVM.CUNY.EDU (Jeff Sicherman,CSU Long Beach) Subject: Mouse book vs. manual Microsoft Press (a.k.a. the propoganda arm of Bill Gates enterprises) has just published a book on mouse programming and menus. Unfortunately it was shrink wrapped when I saw it due to the fact it contains some diskettes in the package, so I want able to examine it. Can anybody tell me how different this is from the Microsoft Mouse Programmers Reference Guide that I already have. I dont want to waste 29.95 (plus tax & license) if it's just a mass-market republishing of the same things. Jeff Sicherman jajz801@calstate.bitnet ------------------------------ Date: Mon, 08 May 89 09:39:17 EDT From: Joe Morris (jcmorris@mitre.arpa) <jcmorris@mwunix.mitre.org> Subject: Passing parameters to batch files In INFO-IBMPC 89:50 Franz-Werner Gergen asks about passing parameters to a batch file when they include an equal sign. He is attempting to pass a parameter of the form "keyword=value" but is finding that "keyword" and "value" are being treated as two separate parameters and the equal sign is dropped. I squawked this to IBM some time back; the response I got was "working as designed" (this is a BAD response: it should be "Broken As Designed"). The IBM response points to the IBM DOS manual, page 7-11 where it says: Commands and parameters *must* be separated by delimiters (space, comma, semicolon, equal sign, or the tab key). The delimiters can be different within one command. For example, you could type: copy oldfile.rel;newfile.rel rename,thisfile thatfile In other words, you can't pass the delimiter characters across the batch processor interface. I've tried to figure out a way around the restriction (doubling the characters, quoting them, etc.) with no luck. Going to non-IBM procedure languages (e.g., Personal REXX from Mansfield Software Group) would seem to be the only way to get around this problem, unless someone can suggest a way to convince uSoft to redesign the batch file parsing. Joe Morris ------------------------------ Date: Tue, 9 May 89 12:17 EDT From: psrc@pegasus.att.com Subject: PK's ZIP program Phil Katz's PKZIP doesn't read .ARC files for one simple reason: his agreement with Software Enhancement Associates forbids him from ever writing software that can read or write such files. SEA now says they'll give you a license to distribute such software, just for the asking . . . as long as you give them a copy of the source code, so (they say) they can make sure you aren't passing around anything that trashes ARC files, which would hurt them commercially. Paul S. R. Chisholm, AT&T Bell Laboratories att!pegasus!psrc, psrc@pegasus.att.com, AT&T Mail !psrchisholm I'm not speaking for the company, I'm just speaking my mind. ------------------------------ Date: Mon, 8 May 89 08:55 MST From: GORDON_A%CUBLDR%VAXF.COLORADO.EDU@CUNYVM.CUNY.EDU Subject: printer problems Does anyone have any suggestions about the following problem: I have an AT clone with a Phoenix BIOS, 1 MB ram, a CGA, 2 floppies and a HD. Recently the printer started acting strange. Most of the alphanumeric characters can print either with print screen, transparent print or output from a program. However, none of the control codes, such as <CR><LF> are interpreted so that the output so not get formatted. Further, during a print screen, spaces are output as a character that appears to be an apostrophe. I changed out the parallel port and the printer and got the same results. Any suggestions about which chip(s) on the motherboard are flaky? Thanks Allen Gordon ------------------------------ Date: Mon, 8 May 89 08:29:49 CDT From: moore@ncsc.navy.mil (Moore) Subject: Toshiba 1.44MB Floppy Question This is probably a dumb question, but I'm stuck: I installed a Toshiba 1.44MB floppy in my Zenith 248 at home (running DOS 3.3+, 2.0F ROM BIOS). I've told both SETUP and DSKSETUP that I have a 1.44MB 3.5" floppy. It works fine in 720K format, but here's the problem: If I format a (yes, I know you're not supposed to) 720K disk in my SLT/286's 1.44MB drive as 1.44MB and then copy (say) the Compaq's DOS directory to the floppy, if I then take it to the Zenith, I get bad data errors trying to read it. Also, the Zenith drive makes horrible noises when I try to make it format to 1.44M. Here comes the stupid question: does the Toshiba drive detect the absence of the high density cutout on the 720K disk I'm trying to use and so refuse to use it as HD? The Compaq drive obviously doesn't, since it'll format a 720K to 1.44 with no problem. Thanks for any assistance. Please respond directly to me (moore@NCSC.NAVY.MIL) and I'll summarize. ------------------------------ Date: 9 May 89 09:30:40 GMT From: terra@diku.dk (Morten Welinder) Subject: TSR for Periodic execution of a Function Alessandro Russo <ALEX17%IPVIAN.BITNET@CUNYVM.CUNY.EDU> writes: >I want to build a terminate-and-stay-resident program which uses the int >1c calls generated by the system clock (generally the only instruction for >int 1c is iret) to do something periodically. >I test the system clock and then, say every minute, I want to execute some >code. The problem is this: while I'm executing my code, obviously the >clock ticks continue to go, so int 1c is again generated, and the computer >stops. I have tried to fix this by redefining int 1c to iret before >running my code, and then reredefining it to my code before my code itself >ends. This works fine, but....... only for a short time! >... >I suspect that it's a stack problem (I use the stack in my code to save >all registers), but I couldn't fix it. I can give you the code, if >requested. 1. You may use a kind of flag instead of redefining the vector. Checking, setting etc. *must* be done with interrupts off - unless you really know what you are doing! 2. Be carefull what you are doing with your periodic function: Never call Dos, since it is not re-entrant. 3. If you use a high-level language, You'll most likely get stack overflow if you interrupted Dos. Switching to you own stack is a very good idea. Greetings from Morten 'Sally' Welinder ------------------------------ Date: Sun, 7 May 89 23:07:38 EDT From: Bruce_Burrell@um.cc.umich.edu Subject: TSR for Periodic execution of a Function If you're not averse to supressing the interrupts while within your routine, you could start with CLI, which turns off all ints but NMI, then end with a STI instruction to turn them back on again. ------------------------------ Date: Mon, 8 May 89 13:01 N From: <NIJKERK%HROEUR5.BITNET@CUNYVM.CUNY.EDU> Subject: TSR proc. Yo, About that problem with the timer int., Ur code should be like this : . flag dw ? ( 1 second ~ 18 ticks) . . new_timeint proc far sti pushf call old_timeint dec flag cmp flag,0 je exit mov flag, time to count down call do_time exit: iret new_timeint endp do_time proc near do ur stuff, don't forget the stack !! do_time endp . . install proc near ....... So u set flag to 60*18 for 1 minute. new_timeint will count down, and when it becomes zero it executes ur code, otherwise it'll terminate. I think this should fix ur problem Hint: buy the book :"Memory resident prog. on the ibm pc" ------------------------------ Date: Mon, 8 May 89 12:26 N From: <OGIGUCHT%HLERUL52.BITNET@CUNYVM.CUNY.EDU> Subject: Turbo Pascal 5.0 Bug reports. First there is a small typo error in the TP5 RTL bug report from Peter Sawatzki (digest V89#44). It is the 5th line of SHORTCUT.DIF. This line reads '; PUSH ES -> PUSH ES' and must of course be replaced by '; PUSH ES -> PUSH DS'. The rest of the file didn't contain errors. Second, are there any more bugs with solutions known about TP5 ? I myself have a problem without an answer: I've tried to install TP5 on an Epson PC-E with a VGA graphics card with a monochrome monitor attached. When I start the Integrated Development Environment it loads verything needed from disk and blanks the screen so you can't see what you do anymore. It is still possible thoug to use the keyboard and to leave the environment. I did check everything without config.sys and autoexece.bat, but the problem stays. The funny thing about this is that Turbo Debugger just works fine on the system, so it's definitely the video card/display combination which folls TP5. Can anyone help me with this ? Thanks in advance, o ____ ____ ____ ____ _____ | e-Mail ogiguchte@hlerul52.bitnet / /____) / \__/ )__/____) / / | phone +31-2522-11809 __/__(_______/ (____/ (______/ (_ | / / | p-Mail Kagertuinen 65 (_/ | 2172 XK Sassenheim ----------------------------------------------| The Netherlands Jeroen W. Pluimers |--------------------------------- Gorlaeus Laboratories | Time goes, you say ? Leiden University | Ah, no! Alas, time stays, we go. The Netherlands | (A. Dobson) ------------------------------ Date: Tue, 9 May 89 12:09 EDT From: psrc@pegasus.att.com Subject: Windows 386 problem David, some early versions of the 386 (20 MHz only, I think) had problems when multitasking, using a 387, and doing something else. The right people to call about this are at Intel, since they make both the chip and the board. Before you call, open up your PC, and write down any information you can find printed on the 386 chip. Good luck. Paul S. R. Chisholm, AT&T Bell Laboratories att!pegasus!psrc, psrc@pegasus.att.com, AT&T Mail !psrchisholm I'm not speaking for the company, I'm just speaking my mind. ------------------------------ End of Info-IBMPC Digest ************************ -------