Info-IBMPC@C.ISI.EDU.UUCP (03/28/87)
Info-IBMPC Digest Friday, 27 March 1987 Volume 6 : Issue 19 This Week's Editor: Billy Brackenridge Today's Topics: Make and Break Key Codes Morse Code Cursor Graphics CMU Offers Graphics Package Disk Fixing KWIC Generator Statistical Packages Complex Arithmetic with MS Fortran RAMPAGE Card and 80286 Accelerator Cards Pop Up Scinetific Calculator Talltree JRAM-3 and Jlaser Card Deleting long Directory Path (3 Msgs) ZAP VT100 Tektronix Terminal Emulator Tektronix Emulator via SIMTEL20 SIMTEL20 File Server 3d Animation Program now Available from SIMTEL20 Memory Resident Spreadsheet InstaCalc Available on SIMTEL20 Line Printer Capture Program LPTx v6.00 Parameter Ram Backup More MS Fortran 4.0 FORTRAN Benchmarks Framework Questions Responses Getting Information about MS-DOS Machines Though Networks More Decnet Dos Woes Indentation of Switch Statements Multiple Marks CXTND.E Warning about Mixing EGA and Videodisk Proceedings Editor Shareware and National Characters VAXmate Technical Reference Today's Queries: NETBIOS Information Wanted MicroEmacs Sources Needed ANSI.SYS IBM Token Ring Novell Netware Async Server Problem Bug in MS-DOS Floppy Handling How does DOS Manage its Buffers ARC needed for Unix Framework II Screen Driver for WY700 Expert System Shells Mailing Label Programs Wanted Russian Word Processor Wanted Making Rainbow100 Compatible MS-DOS Goes to Cleveland Digitized Voice 327x Emulators Everex HD/FD Controller Query Secured PC Query Formula to Relate Number of Wait States Memory and Processor Speed DOS 3.2 on Compatibles Speller Dictionaries Creating a New DOS Shell Cross Assembler for 6502 VMS/Vax as Mac PC file server Slowing Autorepeat on Keyboard dDBASE III Frequency Analysis Wanted Microport Unix vs. Xenix Graphics Software Request Integration of PC and VAX Want to Kill Keyboard Repeat Computer Products United Micros Postscript Filters Finding Current Application Name FileCard 20 on AT ---------------------------------------------------------------------- INFO-IBMPC BBS Phone Numbers: (213)827-2635 (213)827-2515 Date: Thu, 19 Mar 87 00:23:32 EST From: "Keith F. Lynch" <KFL%MX.LCS.MIT.EDU@MC.LCS.MIT.EDU> Subject: Make and Break Key Codes Morse Code Thanks to all those who replied to my query regarding make and break keyboard codes in Turbo Pascal. The trick is simply to look at "port[96]" which will have a value less than 128 if a key is down and more than 128 if no key is down. Which value depends on which key is down (or was most recently pressed if more than one key is down) or was most recently released. The down and up codes for a given key differ by exactly 128, and can be found for each key by writing a short test program. For instance on my machine the left shift key gives 42 when down and 170 when released. Given this, it is easy to write a program to for instance make a different musical note depending on which key is down. My application is to write a program to allow the user to "type" in Morse code on any one key, and have the machine know what is being typed. ...Keith ------------------------------ Date: 19 Mar 1987 09:08-EST Sender: MHARRIS@G.BBN.COM Subject: Cursor Graphics From: MHARRIS@G.BBN.COM In reply to Mr. Gerhson's request: you might try Herm Fischer at Mark Five Business Systems, (818) 995-7671. He's developed a screen graphics package which may be useful to you. -- Michael Harris ------------------------------ Date: Thursday, 19 March 1987 17:18:35 EST From: Peter.Levin@henry.ece.cmu.edu To: Info-IBMPC@c.isi.edu Subject: CMU Offers Graphics Package We have just completed a simple MicroSoft C program that easily implements primitive graphics functions on the IBM PC with an Enhanced Graphics Adapter. Additionally, the program also includes an interface to MicroSoft or Mouse Systems mice. The only system requirement is 128K (or more) of graphics memory, not an unusual configuration for people with EGA or compatible boards. The program is extremely well documented. A user's manual is included. It was originally developed as part of another program called the Charge Simulation Program, an electrostatics problem solver. We have simply taken the appropriate procedures and made a stand alone program out of them. Please direct inquiries to : InstructSystems University Technology Development Center 4516 Henry Street Pittsburgh, PA 15213 (412 - 672 - 8030) or send electronic mail to : pete@henry.ece.cmu.edu.arpa ------------------------------ Date: Thu, 19 Mar 87 21:42:11 est From: Mike Ciaraldi <ciaraldi@rochester.arpa> Subject: Disk Fixing In response to Roger Jagoda's request for disk-fixing programs, I would like to report that I have had good results with the "Ultra Utilities". The last version I used was dated 1984, and I don't know if there has been an update since then. The suggested shareware price is about $40. These programs let you view and modify any sector on a floppy, copy sectors between or within disks, and do custom formatting, with different numbers and sizes of sectors. Mike Ciaraldi ARPA: ciaraldi@rochester.edu uucp: seismo!rochester!ciaraldi ------------------------------ Date: Thu, 19 Mar 87 21:49:48 est From: Mike Ciaraldi <ciaraldi@rochester.arpa> Subject: KWIC Generator I'm looking for a program that will generate KWIC (Key Word In Context) indexes (indices?). Preferably public domain or inexpensive. I looked in the various ARPA archives, and the nearest I could find was one for CP/M and CP/M-86 (SIG/M Volume 173). A recent INFO-IBMPC digest featured a KWIC index as part of an answer, so there must be such software around somewhere. Thanks. Mike Ciaraldi ARPA: ciaraldi@rochester.edu uucp: seismo!rochester!ciaraldi [Read Software Tools in (Pascal/C) by Kernighan and Plauger. Kwic generators and Sort routines are a major part of this book. The actual kwic program is about a dozen lines long. If you already have a sort routine available you are 90% done. I use Unix sort. Even with FTP time a real VAX wins when sorting multi megabyte files. -wab] ------------------------------ Date: Fri, 20 Mar 87 11:15 N From: <DRAYER%HGRRUG5.BITNET@wiscvm.wisc.edu> Subject: Statistical Packages Hello all! Sorry...I bungled things up somewhere along the line, and my summary of the replies on my query concerning statistical packages got sent to a black hole somewhere. I hope no-one was too disappointed. So here's what I've concluded concerning the following packages for the PC: SAS-PC: At least 3 times more expensive in Europe than in the US, also for universities, which means this is the most expensive package. This pricing is enough to put me off. And it probably is the slowest of the 4. Takes up a lot of room on your HD...include that in the price too! STATGRAPHICS: Has beautiful graphics. Same price range as Systat. Cannot handle more than a few variables and cases, so it's not really comparable with the other 3. SPSS/PC+: Like SAS-PC, this package takes up a lot of HD room, so that has to be added to the price. Our university has a site license for the 3 main parts of the package, so it's not too expensive. Taking the extra needed HD room into account, the package is slightly more expensive than Systat or Statgraphics. The package can be classed as "semi- interactive" I'd say. It's fast when a coprocessor is available. The commands look like, but are not the same, as SPSS/X. Why o why did they have to make them different? No programming language like in SAS or Systat is available. SYSTAT: To get to the point, this is the chosen one... It's the only package of the 4 to include time series analysis. The main drawback of Systat is it's very limited graphical capabilities. It's fast, takes up less than a third of the HD space needed by SPSS/PC+, and can even be run from floppies. Two versions are available, the extended version utilizes 512K RAM, the other less. Can anyone tell me WHY WHY WHY people in the US pay $595 and we in Europe have to pay $695 (yes, an extra $100). I like Systats language, it's much terser than SPSS or SAS. This is the main reason I chose this package. I hate remembering and typing long commands. I hope this was not too chaotic. This wasn't intended as a product review. Other products were mentioned by some people (GAUSS). It seems to me however that the above 4 are most used and known. Nick Drayer jr Groningen University Holland Bitnet: DRAYER@HGRRUG5 ------------------------------ Date: Fri, 20 Mar 87 10:49:18 EST From: Bernard_Tiffany@um.cc.umich.edu Subject: Complex Arithmetic with MS Fortran Regarding to the topic "Complex Arithmetic with MS Fortran" on Info-IBMPC digest, Volume 6: Issue 17. > I have discovered that the Microsoft Fortran compiler v. 3.31 > produces very poor code for complex arithmetic ... [especially] > complex multiplications. It is true that mainframe Fortran compilers such as IBM VS Fortran compiler V2.1.1 generate inline code for complex multiplications for speed. For example, on an IBM 3090-400 computer, 1000 complex multiplications took 938 microseconds with OPT(3). The vectorized code is even better: 436 microseconds with OPT(3) VECTOR. Microsoft Fortran Optimizing compiler v4.00 does not generate inline code for complex multiplications; the subroutine __FHcfmulcd was called along with several 8087 instructions (two fld's and two fstp's). With Microsoft Fortran Optimizing compiler, I tried the complex multiplications on 5-element COMPLEX*16 arrays 1000 times. The complex multiplications took 1.26 seconds on an IBM PC-XT with 8087 and 640K memory; and the scalar equivalent took 1.76 seconds. The default is now $NOFLOATCALLS with 8087 library. Bernard Tiffany University of Michigan Ann Arbor, Michigan ------------------------------ Date: Fri 20 Mar 87 12:52:55-PST From: Ivan Auger <LAWRENCE.Auger@BIONET-20.ARPA> Subject: RAMPAGE Card and 80286 Accelerator Cards I have solved the problem of using an AST RAMPAGE card on an IBMPC/XT with a 286 Express accelerator card (which does caching). The trick is to load in a device driver to trap parity errors. This should be the first driver. Then you should load in the REMM.SYS driver. ------------------------------ Date: 20 Mar 87 17:09:00 EST From: "V703::S_DANIELS" <s_daniels%v703.decnet@nusc> Subject: Pop Up Scinetific Calculator Some people have been asking about this. I put up a msg some time back about 2 versions: - the commercial version - my own kludge The commercial version is PRO/SCI and PRO/BIZ, from SYMSOFT. Phone is 415-962-9500. I don't know if they are still in business. Another company has one called KSH-1 SCIenmtific Calculator. This is from K SOFTWARE HOUSE in Tennessee. Phone 615-294-5090. Was reviewed in PC Tech Journal July 86. My own effort grew out of Turbo, Pascal's MINI CALC spreadsheet. Since I have source code, it is easily extensible. I haven't worked on it in a while. If anyone is interested, call me weekdays and we can arrange a transfer. Scott Daniels (203) 440-5327 Naval U/W Systems ------------------------------ Date: 20 Mar 87 14:35:30 PST (Friday) Subject: Talltree JRAM-3 and Jlaser Card From: "BOB_FORSYTH.XRCC-NS"@Xerox.COM I have been struggling with a Talltree JLASER card attempting to obtain some reasonable output from our Cordata printer. Running the Joutput program supplied I get the described cross/box but the left edge of the print comes out black for about .5 inch, it has a firm left edge from the start of scan and a ragged right edge. The raggedness carries across to make the lines that are intended by the program fuzzy. I have also tried to output a picture file from Ventura Publisher and have obtained the same black band and raggedness in the picture. I have tried to talk to Talltree and have tried out their various changes to the driver defaults with no improvement. This note comes after inserting the third replacement board, a solution they usually arrive at when all their instructed changes to driver default fail. The problem appears to me to be some kind of timing error that occurs while transferring the data file to the printer. I hope that someone out there may have found a solution that can be passed on to me and any others who may be struggling to end this trouble. BOB ------------------------------ To: Info-IBMPC@C.ISI.EDU Subject: Deleting long Directory Path From: marcl@3Com.COM (Marc Lavine at 3Com Corporation) Date: Fri, 20 Mar 87 21:52:08 PST In Volume 6, #18, Gene Hart asked about how to remove a directory with the name: \junk1\junk2\junk3\junk4\junk5\junk6\junk7\junk8\junk9\junk10\junk11\junk12 He said that he could cd all the way down to junk10. To get rid of the lower directories from there, just try "rmdir junk11\junk12" and then "rmdir junk11". What I've found is that DOS's limit of 64 characters for directory names is only a limit for CHANGING to that directory. DOS allows you (programmatically, anyway) to create and otherwise manipulate directories (and files in them) with significantly longer names. In fact, if you had a program that had arbitrarily large buffers for directory names, I suspect that you would find that there is NO length limit for most operations (since I don't think DOS needs to store the name of files (other than an executing batch file, perhaps) to manipulate them, but simply accesses them by traversing the directories in the name, no matter how many levels there are). Of course, these directories aren't that great since you can't "cd" to them and they screw up normal programs expecting 64 character maximum paths that try to reference them. ------------------------------ Date: 19 Mar 87 3:05 -0800 From: "Ya`akov N. Miles" <bd%dac.triumf.cdn%ubc.csnet@RELAY.CS.NET> Subject: Deleting long Directory Path In order to delete your "\JUNK\JUNK\JUNK\JUNK..." directory, may I suggest that you run one of the many disk repair utilities, and modify the root directory "\JUNK" so as it loses its "directory" status, then delete it. You will have a bunch of lost clusters - recover them with CHKDSK /F and then delete all files in the root directory ending in CHK reply: <bd@triumfcl.bitnet> ------------------------------ Date: Thu, 19 Mar 87 15:52:07 CST From: Jess Anderson <anderson@unix.macc.wisc.edu> Subject: Deleting long Directory Path Digest vol. 18 had a query about how to delete an oddball directory. Recently I (foolishly :-) tried a CGA simulator program obtained from a local BBS, and upon losing interest in trying to get it to go, I discovered a new directory on my hard disk (could have been worse, right?). It had a one-byte name, FE (hex), which I couldn't figure out how to enter as a parameter for the rmdir command. However, an excellent file manager program called QFiler, available on many BBSs and at SIMTEL20, lets you point at the thing, mark it, and delete it, obviating the necessity of entering the name. Worked like a champ. QFiler is shareware. Standard disclaimers, etc. Jess Anderson (anderson@unix.macc.wisc.edu) ------------------------------ Date: Fri, 27 Mar 87 12:03:47 EST From: swatsun!byzantium!platt@seismo.CSS.GOV (Steve Platt) Subject: ZAP VT100 Tektronix Terminal Emulator The address for obtaining ZAP (terminal emulator/VT100/Tektronix) is probably incorrect, although the post office may forward it. Paul Fishwick, author of ZAP, moved to Florida last year. I *have* seen ZAP advertised in various periodicals (Dr. Dobb's, for instance) by one or more of the standard mail-order places. I believe Solution Systems was/is marketing ZAP. -steve platt ------------------------------ Date: Thu, 19 Mar 87 19:17:26 CET From: Eberhard W. Lisse <LISSE%DACTH51.BITNET@wiscvm.wisc.edu> Subject: Tektronix Emulator via SIMTEL20 Hi, Pete Galvin has responded to my last message. I'll send the package to him and he will install it on SIMTEL20.ARPA. I do not want to send the same file more than once over the gateway if it is not necessary. He will be posting a message once he has it on line. I'll send it direct to BITNET users or users without FTP access to SIMTEL20.ARPA. el [There were several messages one from London and the other from Australia asking that the terminal emulator be sent as mail as file access is not practical over long distances. Those people and others please note SIMTEL20 can now send you these binary files as mail! Please don't ask info-ibmpc we will just direct you to SIMTEL20. -wab] ------------------------------ Date: Sun, 22 Mar 1987 07:08 MST Sender: KPETERSEN@SIMTEL20.ARPA From: Keith Petersen <W8SDZ@SIMTEL20.ARPA> Subject: SIMTEL20 File Server > Pardon my ignorance Keith, but what is the user name of your file server > for net mail? If I would like to fetch a file using net mail I should > state the filename in the subject line, shouldn't I? I guess you get a > lot of letters asking for the same thing all the time, but I hope you > don't mind helping me... I don't mind at all. The files are here for everyone to access. Internet users should use FTP rather than net mail. FTP transfers the files much faster and avoids placing additional load on the mailer. The subject line is ignored by the file server. The file server cannot answer requests which use a path that includes seismo. The most successful route is ucbvax because they have an excellent program that properly formats the "From:" line of messages coming through the gateway. To obtain up to five files in a single request message by net mail from the public domain archives kept on SIMTEL20.ARPA, send a message to: ARCHIVE-REQUEST@SIMTEL20.ARPA or via uucp: ...!ucbvax!simtel20.arpa!archive-request ...!uw-beaver!simtel20.arpa!archive-request ...!decwrl!simtel20.arpa!archive-request ...!lll-crg!simtel20.arpa!archive-request ...!ut-sally!simtel20.arpa!archive-request ...!harvard!simtel20.arpa!archive-request [do NOT use host "seismo" - they are blocking messages from the server] The message body must contain lines beginning with the keyword SEND, one SEND line for each file requested. Case is not significant. The general syntax of a SEND line is: SEND format filename In general, a filename consists of the following components: device:<directory>file.type.generation "device:" is usually PD:, and the combination of PD:<directory> is expected unless an alias has been advertised of the form "alias:", which takes the place of both device and directory fields. The generation field should be left off in order to default to the highest generation number so you can be sure of getting the latest version of the file requested. "file.type" follows the usual file naming conventions. In all formats listed below, if the file to be sent is larger than 55K, the file is sent in numbered parts. The parts must be reassembled in order and edited to remove any headers, preface, and trailers before the process can be reversed to reconstruct the original file. Allowable formats are: SEND HELP This file you are reading now. SEND INFO A detailed description of the SIMTEL20 Archives, which includes this file, pointers to certain key files, and descriptions of various file transfer programs and related utilities. SEND BOOTSTRAP A brief quick reference listing of filenames of the key utilities used to reconstruct files sent by the compression and encoding techniques listed below. SEND DIR filespec This format returns a CRC list of the requested files, and is the only format which allows wild card filenames (but not wild card directory names). The list is sent as an ASCII text file. The wild card characters are "*" and "%". The asterisk means any number of characters, while the percent sign means exactly one character. Either or both may appear in any combination in either or both the file or type fields, while only the asterisk may appear in the generation field. SEND RAW filename If the file is ASCII, it is sent as-is, regardless of size. This format is the least efficient over network and mail gateway resources. Use this format only if you absolutely must. With the four formats listed below, if the file is ASCII and under 25k characters, it is sent as-is, as if RAW format was requested. Binary files are always processed according to the requested format. However, a request for ARC or SQ processing of files with type ".ARC", ".LBR", or ".%Q%" is ignored and the original file is either uuencoded or hexified (if possible), according to the requested format. If the file was not sent RAW, a short preface is inserted at the front of the message describing the process actually taken and a CRC entry describing the original file. SEND ARE filename or SEND filename The original file is made into a uuencoded ARC file. SEND ARH filename The original file is made into a hexified ARC file if the ARC file is under 64K bytes long. Otherwise, an apology is returned instead of the requested file. SEND SQE filename The original file is made into a uuencoded SQueezed file. SEND SQH filename The original file is made into a hexified SQueezed file if the Squeezed file is under 64K bytes long. Otherwise, an apology is returned instead of the requested file. To get started in finding your way around the SIMTEL20 archives, send a message to the server with the request: SEND INFO --Keith Petersen Arpa: W8SDZ@SIMTEL20.ARPA Uucp: {bellcore,decwrl,harvard,lll-crg,ucbvax,uw-beaver}!simtel20.arpa!w8sdz GEnie Mail: W8SDZ [OK Now all you folks world over who want shareware sent to you as mail here is your chance! -wab] ------------------------------ Date: 12 Mar 87 16:45:25 GMT From: dalcs!dalegass@SEISMO.CSS.GOV (Dale Gass) Subject: 3d Animation Program now Available from SIMTEL20 Now available from SIMTEL20... Filename Type Bytes CRC Directory PD:<MSDOS.SCREEN> 3DM.ARC.1 BINARY 13109 AD49H 3DM is a 3d animation program for the IBM-PC with CGA. I posted a similar one about a month ago. This one is much more flexible, and uses '.mov' files to find out how to set up the windows, what figures to use, and how to rotate them. The file 'descript.mov' describes the basic layout of these files. There are 5 others that can actually be used: 3d.mov - a group of 5 objects - similar to the original prog. cube.mov - a single cube cyl.mov - a cylinder sur.mov - a 3d surface sphere.mov - a sphere By passing these file name to the '3dm' program, it will interpret the files: i.e. '3dm cyl' will draw a large cylinder on the screen, as described in the cyl.mov file... If you have limited memory, you may need to modify the step-size parameter in the '.mov' files to make coarser animation. Enjoy... *************************************************************************** Dale Gass, Dalhousie U., Halifax, N.S., Canada UUCP: {seismo|watmath|utai|garfield}!dalcs!dalegass.UUCP or dalegass@dalcs.UUCP ...!dalcs!dalcsug!dalegass.UUCP or dalegass@dalcsug.UUCP CDN: dalegass@cs.dal.cdn CSNET: dalegass%cs.dal.cdn@ubc.csnet ARPA: dalegass%cs.dal.cdn%ubc.csnet@CSNET-RELAY.ARPA ------------------------------ Date: Thu 19 Mar 87 08:13:43-PST From: Bruce Buzbee <BUZ@SRI-KL.ARPA> Subject: Memory Resident Spreadsheet InstaCalc Available on SIMTEL20 I have transferred over to SIMTEL20 an archived copy of InstaCalc, a user supported memory resident spreadsheet. Filename Type Bytes CRC Directory PD:<MSDOS.SPREADSHEET> INSTA.ARC.1 BINARY 100516 11A0H Program: InstaCalc Company: The FormalWare Co. P.O. Box 21726 San Jose, CA. 95151-1726 Registration: $35 (+$5s/h) Description: InstaCalc is a fast, full feature memory resident spreadsheet. It has 256 columns and 256 rows. - You can pop up in the middle of an application (word processor, data base, etc) and import and export data between InstaCalc and the application. - You can directly import Lotus .WKS files - 44 pre defined functions - context sensitive help screens - powerful macro facility - can confine recalculation to a range of cells - hidden columns - You can perform any kind of modification on all the numbers in a range. For example, you can have InstaCalc double all of the numbers in a range. - And a whole lot more... ------------------------------ Date: 19 Mar 87 06:17:23 GMT From: esosun!ucsdhub!jack!man!wolf!markd@SEISMO.CSS.GOV (Mark Divecchio) Subject: Line Printer Capture Program LPTx v6.00 Organization: Systems'n'Software San Diego, CA Now available from SIMTEL20... Filename Type Bytes CRC Directory PD:<MSDOS.PRINTER> LPT600.ARC.1 BINARY 31596 CCB2H LPT600.ARC is version 6.00 of my Line Printer Capture program LPTx. LPTx allows you to redirect most printer output to a disk file. This version works fairly well under DOS 3.x which caused earlier versions to fail. I have been distributing version 6 as a Beta release for a few months and users have had enough success to warrant this posting. The source and documentation describe the conditions under which it works and under which it does not work well. Enclosed in the ARC is the source in assembly language, a documentation file and the .COM executable. Anyone interested in the earlier versions of LPTx or on the lookout for future updates can check with the Fido Bulletin Board at 619-741-3412. Mark C. DiVecchio 10435 Mountain Glen Terrace, San Diego, CA 92131 K3FWT Home of PC-VT and LPTx 619-566-6810 sdcsvax!man!wolf!markd ------------------------------ From: Ralf.Brown@b.gp.cs.cmu.edu Subject: Parameter Ram Backup Date: 15 Mar 87 19:49:34 GMT Organization: Carnegie-Mellon University, CS/RI I too have an AT clone with 4 pen light batteries for the clock. Here are two small BASIC programs I hacked up a few months ago: the first reads the NV-RAM and displays its contents on the screen, and the second restores the contents from DATA statements in the program (which you would put in according to the values displayed by the first program). The first program displays three columns of three numbers on the screen. For each group of three numbers, the first is the address (in hex), the second and third are the value in hex and in decimal. 10 FOR X = 0 TO 63 20 OUT &H70,X+128 30 Y=INP( &H71 ) 40 IF X < 16 THEN PRINT "0"; 50 PRINT HEX$(X);" "; 60 IF Y < 16 THEN PRINT "0"; 70 PRINT HEX$(Y);" "; 80 PRINT RIGHT$(" "+STR$(Y),3); 90 IF X MOD 3 = 2 THEN PRINT ELSE PRINT ,, 100 NEXT 110 END 1000 CLS 1010 FOR X = 1 TO 2500:LOCATE 1,1 1020 OUT &H70,128 1030 Y=INP( &H71 ) 1060 IF Y < 16 THEN PRINT "0"; 1070 PRINT HEX$(Y);" "; 1080 PRINT RIGHT$(" "+STR$(Y),3); 1090 NEXT Change the DATA statement in the following to contain the values displayed by the previous program for locations 0E to 32 (hex) inclusive. You will need to set the time and date after restoring the NV-RAM (unless you want to modify the second program). I didn't include the locations containing time and date since they obviously change while the rest of the RAM doesn't unless you run the setup program. 10 FOR X = 14 TO 50: READ Y: OUT &H70,X:OUT &H71,Y:NEXT 20 DATA 0,0,16,0,240,0,49,128,2,128,1,17,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 21 DATA 0,0,2,69,128,1,25 +=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-+ | ARPA: RALF@B.GP.CS.CMU.EDU "Teaching COBOL ought to be | | AT&T: (412) 268-3053 (school) regarded as a criminal act" | | Snail: Ralf Brown --- Edsger Dijkstra | | Computer Science Department | | Carnegie-Mellon University DISCLAIMER? Who ever said I claimed| | Pittsburgh, PA 15213 anything? | +=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-+ ------------------------------ Date: Mon, 23 Mar 1987 14:42 PST From: JAJZ801%CALSTATE.BITNET@wiscvm.wisc.edu Subject: More MS Fortran 4.0 To: info-ibmpc@c.isi.edu I submitted the original query about 'what do I get for $150 from MS Fortran 4.0'. The replies were not very enlightening other than mentioning Codeview but I guess this is because there is not really much to say other than that it brings MS Fortran (finally) up to ANSI 77. I still would like to hear about conversion problems from anybody with any experience with it. Interestingly, the campus bookstore here is offering the new version for $170. I don't know how close to cost that is and whether it represents educational discounts but that is so close to the $150 upgrade to indicate that the wholesale price and the upgrade cost are nearly equal. It suggests to me that MS (and perhaps the other vendors) see the market for some development tools as relatively saturated and that the only way maintain revenues is to hook them low and then price the upgrades higher once you've got them to keep the cash flowing because you won't get many more new ones. That, and the assumption that us developers are rich enough to buy expensive software often (not always true for educational institutions guys). Though I don't expect it, I really would like to hear what somebody from Microsoft has to say about all this. Jeff Sicherman ------------------------------ Date: Mon, 23 Mar 87 12:24:11 +0100 (Central European Time) From: XBR4D715%DDATHD21.BITNET@wiscvm.wisc.edu Subject: FORTRAN Benchmarks Hi Billy! Some time ago I made some measurements with various compilers/ machines. Microsoft FORTRAN is quite slow compared with Ryan McFarlands F77 (full Fortran 77) The first test was double prec WHETSTONE: (kilo-Instru's/sec) -------------------------------------------------------------- >>> DOUBLE PRECISION WHETSTONES <<< -------------------------------------------------------------- machine I compiler I -------------------------------------------------------------- I MS V3.31 I RM V1.01 I RM V2.10 I I--------------------------------------------------- I ----------- options ------------------- I I--------------------------------------------------- I float I no_fl. I /b I I 80286/287 I I--------------------------------------------------- AT 6MHz I 77 I 73 I 82 I 106 I 109 I AT 12MHz I 143 I 136 I 141 I 194 I 200 I XT 10MHz I 88 I 77 I 154 I 137 I I ATT 6300 I I I I I I V30, 8MHzI 112 I 104 I 166 I 173 I I -------------------------------------------------------------- To compare: VAX 11/780, FPA, VMS 4.4 : 800 KIPS VAX 8500, FPA, VMS 4.4 : 2500 KIPS The second test I used was the well known (?) double prec. LU benchmark by Jack Dongarra, counting floating point OP's, not instru's. The results were: -------------------------------------------------------------- machine I kFlops I -------------------------------------------------------------- ATT 6300 8087-2, DOS 2.11, RM V2.10 I 29.7 I VAX 11/750 w/o FPA, VMS 4.4 opt. comp. I 30.5 I VAX 11/780 with FPA, VMS 4.4 opt. comp. I 151.9 I VAX 8500 with FPA, VMS 4.4 opt. comp. I 565.9 I -------------------------------------------------------------- These measurements lead to the conclusion NOT to use MS Fortran. RM Fortran can be linked with MS FORTRAN/PASCAL/C routines, so there are no problems with existing LIB's. Hope I was of help ! If you are interested, I can send You the sources of the programs used ! Greetings from EUROPE, Klaus D. Schmitt Institute f. El Power Supply Technical University Darmstadt Federal Republic of Germany (DATEX-P: 45615130514::BR4::D715) ------------------------------ Date: 23 Mar 87 09:24 EST From: David A. Potter / McDonnell Douglas <DAP.MDC@OFFICE-1.ARPA> Subject: Framework Questions Responses I received ten responses to my query about a place to which Framework-related questions could be sent. Several were offers of assistance (for which I am grateful); all expressed interest in starting a net interest group / mailing list to support Framework-related dialog. Unfortunately, no one is in a position to handle the work associated with such a list.... So I suppose this will go no farther. But I did feel that all those who responded deserved to know the outcome. Thanks -- David [I can understand why nobody would volunteer to run a mailing list! -wab] ------------------------------ Date: Tue, 24 Mar 87 16:53:24 MEZ From: IWI7B3%DERRZE1.BITNET@wiscvm.wisc.edu Subject: Getting Information about MS-DOS Machines Though Networks * Getting information about MS-DOS machines through wide area networks * Thorsten Glattki <IWI7B3@DERRZE1> There are, as everyone of us knows, file servers for several subjects out in netland. You can find them in EARN, BITNET, ARPANET, UUCP and all other more or less well known wide area networks. They cover several subjects: documents (e.g. for networking like the RFC's and all other formats and protocols), articles and information about several machines and public domain software for some widespread operating systems (most of these files in BITNET are for VM/CMS, and in UUCP most are for UNIX). Working with MS-DOS machines (e.g. IBM-PC and clones) as well as with big mainframes in the above networks one fine day I thought that there should be also some information about these machines in netland reachable using our mainframes. So I started looking around where to get such things. Thanks a lot to all people who gave me tips where they had heard of something similar. During the last 365 days I compiled a list of where to get files for MS-DOS machines. In the meantime some friends of mine in several institutions got to know about this list, requested it and found it useful. As some asked me to offer the list to public, I asked some network administrators of some file servers whether they could install it on their servers. Now the list is available through LFCNET@ICNUCEVM.BITNET, the fine file server in Italy of Niccolo Avico. Perhaps within this year it will be also available through two more servers. But we only have just installed the file list at LFCNET and not all information files In this file list which is called ..... on LFCNET and PC FILE LIST elsewhere, you will find the names of other useful files : help files and file lists (directories) of several servers which are offering files for MS-DOS machines, names, addresses and information about several user groups/mailboxes/mailing lists which are offering more information about MS-DOS and related subjects. Up to now unfortunately you have to request all those help files individually from my own account by just sending me mail that you want to receive all PC FILE LIST files. I will let you know when all other files are placed on public disks. Please note: PC FILE LIST will only tell you where to get more special information and files. I think that PC FILE LIST is quite complete and up-to-date. If you should find other servers or new info about file servers please inform me about what has changed so I can update FILE LIST. Special thanks to Chris Condon's BITNET SERVERS, Gligor Tashlikov's NETWORKS FILE LIST, Richard Zellich's List-of-Lists, Nicco Avico's LFCNET server, Mr Berthold Pasch (one of the men working for NETSERV@DEARN) and -not to forget it- CSNEWS's VM/COM which had been very helpful to compiling my list and then to offering it for public use. Keep up the good work ! ------------------------------ Date: 25 Mar 87 10:20 -0800 From: "Ya`akov N. Miles" <bd%dac.triumf.cdn%ubc.csnet@RELAY.CS.NET> Subject: More Decnet Dos Woes We have installed DECNET-DOS 1.1 on our (generic) AT running PHOENIX bios, with the 3COM interface. Everything goes fine for about two minutes after booting the AT, at which point the AT gets dropped off the (Ethernet) network Subsequent attempts to connect to other nodes using SETHOST cause the AT to CRASH, and the only cure seems to be a re-boot. We have disabled MULTICAST and SERVICE on the circuit. Does anyone have any further ideas? Please reply: <bd@triumfcl.bitnet> <tryn@ubcmtsg.bitnet> ------------------------------ From: <kubitron@ATHENA.MIT.EDU> Subject: Indentation of Switch Statements Multiple Marks CXTND.E Date: Thu, 26 Mar 87 14:25:18 EST This is in response to Jim Anderson's query regarding Epsilon and the indentation of switch statements. Following, is the eel code for and extended C mode, which I have called "C-Extended." This does a reasonable job of indenting switch statements. Furthermore, it adds ten marks which can be set with CTRL-F1 through CTRL-F10. You can return to a previously set mark with ALT-F1 through ALT-F10. Returning to the position from which you last jumped is simply accomplished with ALT-ESC. Obviously, these bindings are not etched in stone--see my comments at the beginning of the file. I cleaned up a few annoying things. On a backward delimiter match (initiated by ']', ')' or '}'), delimiters within quotes (both single and double) or in comments are ignored. Indentation within comments now matches the previous line. This allows comments to be placed at columns other than column 0 without having to fight the auto-indenter. Other random things have been done. Look at the comments at the beginning (after Lagaru's standard c-mode comments), and at the code itself. This code is relatively recent and could be slightly buggy. If you find any problems or have any suggestions, send me a message. I have other things that I plan to do with this code at some point in the ill-defined future. Perhaps I will send an updated version. John Kubiatowicz KUBITRONICS information systems kubitron%athena.mit.edu [CXTND.E can be found in the info-ibmpc .EEL lending library. It along with all the other .EEL code can be found in the <INFO-IBMPC.EEL> directory on C.ISI.EDU. This library isn't on the BBS yet. -wab] ------------------------------ From: gaia!zhahai@seismo.CSS.GOV Date: 26 Mar 87 06:24:18 GMT Subject: Warning about Mixing EGA and Videodisk Proceedings Editor Organization: Gaia Corp., Boulder, CO Regarding gilles' question about a graphics editor for the EGA: there are many editors which use the EGA display as one of their options, indeed almost any commercial paint program (*Paint*) nowadays supports the EGA. If you want to create a screen and then load it later from a C program you will have to check with the paint program vendor for technical info on the image format (on disk), or for utilities/libraries to do this. I hope that you realize that the IBM EGA, in its new screen resolutions, is not compatible with normal video (NTSC in the US) frequencies; it gets its 350 lines of vertical display by increasing the horizontal scan rate far beyond normal video tolerances. Thus you will most likely have a great deal of trouble mixing that output with a normal videodisk (as your note mentioned your application to be). Of course, the EGA also has a 640x200 mode that has sweep rates pretty close to standard, close enough for some equipment to lock (VCR, for example) and with a timebase corrector you could probably mix it with analog video - you still get the EGA's 16 colors, so this might be worthwhile. Of course, the EGA is not interlaced, but some equipment has no trouble with this. Finally note that the IBM EGA does not have a composite (analog) video output like the CGA (the 2 RCA jacks are not connected - they go to an onboard expansion port only) - so unless your mixing equipment handles digital RGBI (or RGBrgb) inputs you may need either an external combining circuit or a clone which does provide composite video. (Composite video contains the picture and sync signals combined into one analog signal; same stuff as a VCR/videodisk puts put with RCA jacks usually, Not the stuff on 9-pin "D" connectors nor the stuff on type "F" "antenna" jacks). You may already understand all this, but it is a source of much confusion for many people. One other route, which I probably would not recommend but would try if I had to, would be to use a good monitor and camera to "scan convert" the EGA higher resolution output to normal video. Besides noise, fuzziness, and distortion, this might make it harder to detect the "transparent color" where analog video should be substituted, depending upon your equipment. To be clear: point a normal NTSC camera at an EGA compatible monitor. (NTSC is the video standard used in the US and Japan for analog video). (Actually, for the color modulation therein, but close enough for here/now). I think the seminar proceedings you are referring to is "IBM Personal Computer Seminar Proceedings", Vol 2, number 11, Nov. 84, which covered the EGA. Please note that this document is incomplete, not very coherent, and sometimes incorrect - it is far from a good "how-to" for the EGA; its principle claim to fame was that it was available to many people before the real technical reference was so it was the ONLY info for a while; it may still have some good info mixed in but not all that much. I don't know where to get it (somewhere in IBM land...), but perhaps the exact title above will help. Zhahai Stewart {hao | nbires}!gaia!zhahai ------------------------------ Date: Thu, 26 Mar 1987 11:38:56 ULG From: Andre PIRARD <A-PIRARD%BLIULG12.BITNET@wiscvm.wisc.edu> Subject: Shareware and National Characters Several programs by private software writers purposefully ignore non U. S. codes from the keyboard. This makes them useless in other countries where one of the reasons for IBM PC success is its multilingual design. May I advise concern about this, in particular to those expecting reward? Also a useful feature is to keep messages in files, so that they can be translated. ------------------------------ Date: 26-Mar-1987 1300 From: waters%rainbo.DEC@decwrl.DEC.COM Subject: VAXmate Technical Reference I have had a large number of requests for VAXmate technical information, particularly in the area of the Ethernet datalink interface. The "VAXmate Technical Reference Manual" (volumes 1 and 2) is currently being printed and will be shipping any day now. The DEC order number is Q6ZCS-GZ. I believe this indicates both volumes. The amount of detail in the technical reference is extensive, complete with programming examples. You may be able to get it through DECdirect (800) 258-1710, or through a local DEC office. I don't know the cost. A summary of the table of contents (the latest that I have seen; this may be different from the actual): VOLUME I Chapter 1 VAXmate Workstation Overview Chapter 2 VAXmate Microprocessor Chapter 3 Interrupt Controllers Chapter 4 DMA Controller Chapter 5 Real-Time Clock and CMOS RAM Chapter 6 Three-Channel Counter and Speaker Chapter 7 Video Controller Chapter 8 Keyboard-Interface Controller and Keyboard Chapter 9 Serial Communications Chapter 10 Mouse Information Chapter 11 Diskette Drive Controller Chapter 12 Hard Disk Drive Controller Chapter 13 Network Hardware Interface VOLUME II Chapter 14 System Startup Chapter 15 ROM BIOS Chapter 16 Programming the VAXmate Under MS-DOS Chapter 17 MS-Windows on the VAXmate Chapter 18 VAXmate Network Software (includes how to interface to Datalink, etc.) Lester Waters Digital Equipment Corporation Personal Computing Systems Group "All of the usual disclaimers apply" ------------------------------ Date: Thu, 19 Mar 87 11:53:49 PST From: rmr@sdcsvax.ucsd.edu (Robert Rother) Subject: NETBIOS Information Wanted Does anyone know where I can get information on NETBIOS, in particular the programming interface. Thanks, Robert Rother rmr@sdcsvax.ucsd.edu [Try the Options and Adapters manual published by IBM. Every site should have a subscription to this important publication. There is also an IBM Technical Seminar proceeding. I suspect additional information is available when purchasing a token net adapter for the PC. Also the mailing list: <pc-token-ring@ucla-ccn.arpa> is probably a better source of information. -wab] ------------------------------ Date: Fri 27 Mar 87 01:16:41-PST From: Jean-Pierre Dumas <DUMAS@SUMEX-AIM.STANFORD.EDU> Subject: MicroEmacs Sources Needed I am used to MicroEmacs v30 on my home ST, to deal with the office PC (really an Olivetti M24, alias AT&T 6300), I would like to have the same but: I am missing the tty sources for PCs. Can anybody provide me with information where to get them missing source files, or send them to me ? I can use MSC 4.0 to compile the beast. I need sources because I hacked up a lot the original v30, with a lot more feature, most notably to support international character set and startup file. Thanks in advance. Jean-Pierre H. Dumas (dumas@sumex-aim.stanford.edu) ------------------------------ Date: Fri, 27 Mar 87 12:48:29 +0200 From: <JOHN%TAUNOS.BITNET@wiscvm.wisc.edu> To: INFO-IBMPC@C.ISI.EDU Subject: ansi.sys I wrote a letter some time ago asking if anybody had any information about the limitations of key redefinition when using ANSI.SYS as I am still at a loss I would very much appreciate help on the subject. Thanks in advance Jonathan <JOHN@TAUNOS> P.S. Some time ago somebody asked a question about Hebrew for the IBM PC I like most people in Israel work in Hebrew as well as in English IBM has a Hebrew National language diskette with special routines for reversing print direction and so on. The characters I'm afraid are defined in ROM. As at the time I didn't have my own code I didn't get round to answering I will be pleased to forward any other information I can to whoever it was that asked. Jonathan [Perhaps the reason you got no response is because ANSI.SYS doesn't have anything to do with keyboard redefinition. Also note again that NANSI.CAT in the info-ibmpc library contains complete sources for an extended ANSI.SYS. You can modify it to do anything you would like. -wab] ------------------------------ Date: 27 Mar 87 12:37:00 EST From: <zn0chapman@nardacva> Subject: IBM Token Ring Novell Netware Async Server Problem To: "pc-token-ring" <pc-token-ring@ucla-ccn.arpa> A question for the folks who use Novell Advanced Netware (and the IBM Token Ring): An add-on product that Novell sells is the Asynchronous Communications Server (ACS). The package consists of the ACS software, a multiport card (4 serial ports/card), ASCOM IV (async comm software), and ATERM (remote dial-in software). The advertised product allows network dial-in and dial-out. A dedicated PC is used as the async server and becomes a "traffic cop". The remote dial-in requires the ACS (as in the dial-out), but also a dedicated PC on the LAN side for program execution; at home you receive screen output and send keyboard input. I've been able to get ACS to work for dial-out to remote async hosts and in a direct connect to our EMail host (VAX). Everything is fine with that (at least). The problem? The remote dial-in feature. I've tried a number of different configurations and I cannot finish a session without the session hanging up. I can exit the dial-in program "gracefully", but the PC doing program execution on the LAN side is locked up. When I connect to the LAN; I appear to be a full functioning member (although screen updates are at 1200 or 2400 bps). Sometimes, a simple DIR will lock the connection up. Some other time, Wordstar might lock me up or dBASE or a Novell MENU or.... The condition CANNOT be duplicated. I have tried: a. Different modems. b. Different ACS servers. c. Different remote execution hosts (IBM PC-AT, Zenith Z- 248). d. Different dial-in systems. e. Different ports on the multi-port card. f. 1200 and 2400 bps operation. I have also: a. At a suggestion from Novell, isolated the three ACS- host-dialin systems from a potential EGA problem (supposedly there is a potential timeout problem with systems and the EGA). All systems had either IBM Monochrome or CGA. Still no luck. b. Re-'gen'ed ANET3 using the fixed ANET3.OBJ file for the token ring (to take care of timeouts in large token rings). Still hangs. During my testing, traffic has been mixed (weekday afternoons (busy) and Saturday mornings from home (no network traffic)). When the session "hangs", the execution host on the LAN side has (usually) one more line of data than the remote PC (e.g. a line was not sent). Does ANYONE know anything about this? Our Novell reps have not been much help. Is there ANYTHING I have NOT tried? Thanks! David Chapman NARDAC Norfolk zn0chapman @ nardacva 804-444-1190 AV 564-1190 ------------------------------ Date: Fri, 27 Mar 87 13:18 EST From: Hess@MIT-MULTICS.ARPA Subject: Bug in MS-DOS Floppy Handling To Gordon L. et. al. on this list: Does anybody know how MS-Word gets around the following MS-DOS bug? (I encountered it with FinalWord II; you can do the same in Epsilon and MS-Windows Write, but somehow MS-Word avoids it!) 1) Write protect a floppy. 2) Open an existing file off it. 3) Save the file back. Get an error: protected. 4) Remove the disk, and unprotect it, and re-insert it. 5) Exit the program. DOS will write to the disk! 6) Run CHKDSK on the floppy. Clusters from your existing file will be considered lost. In other words, in a C program: A: is write protected, ands has a file on it, e.g. A:FOO. fd=creat("A:FOO"); will *not* return a file descriptor, but instead will return an error message. But when you exit the program, it looks like MS-DOS is still attempting to close that (bad) FCB, because it writes to disk. What happens when you run CHKDSK is that you get a zero length file, as if the create really happened, and all the clusters left in the file are on a lost chain. But if you create a file that *didn't* originally exist on the write-protected disk, you'll get the same error code, but the disk won't be written when you exit the program. (1) Why does this happen? (2) Can it be changed for future MS-DOS versions? (3) How does MS-Word get around this problem? Thanks, Brian ------------------------------ Date: Thu, 19 Mar 87 11:31:01 SET From: Alun Saunders <ESC1319%DDAESA10.BITNET@wiscvm.wisc.edu> Subject: How does DOS Manage its Buffers Can anybody out there give me an authoritative answer on how DOS uses the disk buffers as defined in the CONFIG.SYS file ? I wrote a small test program that wrote a file consisting of 'N' 512 byte blocks (the size of a 'buffer' I think ?) and then repeatedly read the file from beginning to end. I had assumed that when the value of 'N' was less than the number of buffers allocated, that subsequent reads to the file would happen with no disk accesses since all the necessary disk blocks would be in memory, i.e no flashing light on the drive. I was wrong ! In fact, the only time this prophecy came true was when the file only consisted of one block ! I had 40 buffers allocated at this time which I would have thought would be plenty. My little test program was written in Turbo, and used the BlockRead and BlockWrite procedures. I traced these calls through and verified that they were indeed using DOS calls and not sneaking into the BIOS. For information, the tests were performed on a vanilla XT with a 10MB disk, running PC-DOS 3.1. ------------------------------ Date: Wed, 25 Mar 87 15:34:54 CST From: moore@ncsc.ARPA (Moore) Subject: ARC needed for Unix I use PKARC on my PC and would like to have a version I can run under Unix. Does anyone know where I can get a copy of ARC that will run under Unix? Jim Moore Moore@NCSC.arpa ------------------------------ Date: Thu 19 Mar 87 09:49:37-PST From: Thomas F. Mandel <MANDEL@SRI-KL.ARPA> Subject: Framework II Screen Driver for WY700 Does anyone out there have -- or know the location of -- a screen driver for Framework II for the high resolution Wyse 700 screen? --Tom Mandel mandel@sri-kl.arpa (after 1Apr87, mandel@kl.sri.com) mandel@well.UUCP ------------------------------ Date: Thu, 19 Mar 87 11:10:41 pst From: Kenzi Karasaki <kenzi@lbl-ux8> Subject: Expert System Shells I need comments on the shells by TI: Personal Consultant (Easy & Plus) Are they suitable for a classification problem? Can they call external routines to process the data they work on? What drawbacks were encountered in real life applications? Thank you for the help. Daniel ps. if you know of any software which compares favorably with these, I'd like to know. ------------------------------ Date: 20 Mar 87 23:56 PST From: William Daul / McDonnell-Douglas / APD-ASD <WBD.MDC@OFFICE-1.ARPA> Subject: Mailing Label Programs Wanted Could someone point me in the direction of any freeware that might be mailing label/IBM-PC-XT oriented? Thanks, --Bi// ------------------------------ From: mit-amt!castillo@mit-amt.media.mit.edu (Brian C. Anderson) Subject: Russian Word Processor Wanted Date: 21 Mar 87 19:53:16 GMT Organization: MIT Media Lab, Cambridge, MA Lines: 17 I would like to know if anyone out there has heard of a word processor for the PC that supports the Russian character set. I am trying to put together a cheap PC system for a friend visiting from the USSR. The system will probably be a clone-PC, monochrome monitor, 256K, 1 serial, 1 parallel, 1 floppy and a printer (probably dot-matrix). The printer will have to support Russian or allow down-loadable character sets. Has anyone seen something like this? Any suggestions would be greatly appreciated. As this is a very esoteric query please reply to me directly, I'll summarize and post later. Thanks, Brian Anderson E-mail: castillo@media-lab.media.mit.edu USnail: 483 Beacon St #52 Boston, Ma. 02115 [Get a Macintosh! This from the editor of info-ibmpc. If you insist on a PC clone there have been discussions of this in past issues. See V5 #20 -wab] ------------------------------ Date: Sun, 22 Mar 87 14:38:28 +0200 From: Oved Ben-Aroya <oved%TECHUNIX.BITNET@wiscvm.wisc.edu> Subject: Making Rainbow100 Compatible Is there a way to make the Rainbow 100 compatible to IBM PC (by adding or changing cards or a black box? )? Please reply by mail to: CCSMDAN@TECHNION.BITNET ------------------------------ Subject: MS-DOS Goes to Cleveland Date: Mon, 23 Mar 87 08:15:25 EST From: jfjr@mitre-bedford.ARPA Here is a problem that I have had no luck with and would appreciate any advice as soon as possible. I am running a spanking new PC/AT 8 mHz, MS-DOS 3.21. I have installed a QUA-TECH data acquisition board. My program to drive this board is written in Microsoft C for the user interface etc. and assembler (Microsoft) for the interrupt handling. It is compiled and linked with the small memory model. The assembler code uses the standard INT 21 calls to get vectors for IRQ3 and IRQ4 and to replace them with my own (I have experimented with other IRQ's but it doesn't help). Now, when I run tests with my own interrupt vectors in place, but when no such interrupts are even produced MS-DOS goes to Cleveland. It won't even respond to control-alt-delete. If I change string literals in the C part of the source, even in places that have NOTHING to do with the interrupts, even in spots of the code that are executed before I borrow the interrupts or after I return them then the program works. The program will also work if I shorten the PATH. When the program works it works fine, I can even give it interrupts and it catches them, logs them etc. but if I go in and modify the prompts (I use printf and fprintf) in random regions of the source then the program stops working and it "dissapears" in places which seem unrelated (just before exiting, just after introducing itself). I have sometimes got it to go into a loop where MS-DOS has trouble reading the COM port ( the program asks a couple of questions) and gives me "Abort,retry or ignore", it somehow can't read any response from the keyboard and I must reboot. If this happens it can be reached with a control-alt-delete. What is going on here?? Why??? How do I stop/fix it?? HELLLPPPPPP!!!!!... Jerry Freedman,Jr jfjr@mitre-bedford.ARPA [Sounds like something is crossing a memory segment boundary. What model C are you using. Does your .MAP file show anything large enough to fill a segment. Just a guess??? -wab] ------------------------------ Date: Mon, 23 Mar 87 10:01:07 est From: Gilles Bloch <bloch-gilles@YALE.ARPA> Subject: Digitized Voice I am looking for a card which records and playback voice, but whose sampling rate is variable and can be set higher than those I found in Info_ibmpc. I'd like to be able to have a 15KHz sampling rate. Does anyone knows of any such card ? [Lloyd Rice of Computalker Associates and the UCLA Phonetics Lab recommends the Adaptive Digital Systems Board for the PC. It has a variable sample rate up to 80KHz. Lloyd also suggests the Frequency Devices 901F variable filter for a Nyquist filter. ADS is in Irvine California and Frequency Devices is in Haverhill Mass. You might be able to contact Lloyd for software that runs with these boards as well as TMS320 software. Computalker Associates is in Santa Monica (213)393-7781 -wab] ------------------------------ Date: Monday, 23 March 1987 12:23pm To: INFO-IBMPC@C.ISI.EDU From: GDAA500@UTA3081.CC.UTEXAS.EDU (OGS.YANCEY) Subject: 327x Emulators Can anyone suggest any 327X emulators for IBM PC's or Compatibles? I am particularly interested in programs that will work with the ITI 3270 boards. Please send replies directly to OGS.YANCEY@UTA3081.BITNET. If someone knows of programs as Freeware or in public domain, all the better. ------------------------------ Date: 23 Mar 86 11:15 EST From: C0144%CSUOHIO.BITNET@WISCVM.WISC.EDU Subject: Everex HD/FD Controller Query I'm pulling my hair out over this one... We've received an Everex EV-390 Controller card that holds both floppy and hard drives. The hard drive portion is working fine. On the floppy side, however, I can't get it to recognize the A> drive by anything other than a continuously lit LED on the drive. We are trying to use only 1 floppy on the system (not 2 as covered by the documentation). Does anyone out there know anything about these beasts? +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ From the North Coast Dave Chatfield, Dept. of Computer Services _____ of America...._-! Cleveland State University ! --___ ___-- ! ! ------(*) ! BITNET: C0144@CSUOHIO ! Cleveland ! ARPA: C0144%CSUOHIO.BITNET@WISCVM.WISC.EDU ! ! USENET: davec!ncoast ! O H I O ! BBS: Assistant Sysop, PC-OHIO 216-381-3320 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ ------------------------------ Date: 23 Mar 1987 16:25-EST Sender: DOWENS@G.BBN.COM Subject: Secured PC Query From: David Owens <dowens@bbng.arpa> Advice or information sought from anyone who owns or operates a PC capable of processing classified data. I need to set up a system and know that it probably entails a removable hard disk and perhaps a special printer. Who sells this stuff at a reasonable price and what else do I need in terms of specialized PC equipment. Thanks in advance, Dave <dowens@bbng.arpa> [IBM Sells Tempest PCs but not at a "reasonable price" -wab] ------------------------------ Date: Tue, 24 Mar 87 10:14:14 SET From: Alun Saunders <ESC1319%DDAESA10.BITNET@wiscvm.wisc.edu> Subject: Formula to Relate Number of Wait States Memory and Processor Speed Does anybody know of a magic formula by which one can easily determine what speed RAM chips one needs for a particular processor (i.e. 8088, 8086, 80286), clock speed, and wait-state combination. I ask this because I have recently purchased an AT-clone which runs at 10MHz, and seems to run with 150ns RAMS and 1 wait-state. What speed chips do I need if I want to run with 0 wait-states ? One source says I need 70ns RAMS, which seems a bit OTT to me ! Alun Saunders ------------------------------ From: <evwong@ATHENA.MIT.EDU> Subject: DOS 3.2 on Compatibles Date: Tue, 24 Mar 87 12:48:17 EST Hello all. This question may have come up and have been answered before but I am new to Info-IBMPC so I'll ask it anyway and you can either refer me to an old volume or answer it. My problem is as such: I am using an old Columbia PC-compatible and want to upgrade to DOS 3.x mainly to use (E)EMS and smaller HD cluster size. I have heard that PC-DOS 3.2 does NOT work on many compatibles and I was just wondering if MS-DOS 3.2 had the same problem. I would assume not, but heck, you never know. Thanks muchly (in advance)! eric ------------------------------ Date: Wed, 25 Mar 87 10:22:40 PST From: jperry@sri-unix.ARPA (John Perry) Subject: Speller Dictionaries Does anyone out there know of any Spelling Checker product with a non-trivial sized dictionary i.e. 10000 words instead of 1000. If so, what is its name, how much does it cost, and where can it be purchased? Thanks to anyone who can answer this query? John Perry ------------------------------ Date: 25 Mar 87 7:05 -0800 From: Jamie Wyatt <buwyatt%brock.cdn%ubc.csnet@RELAY.CS.NET> Subject: Creating a New DOS Shell I am working on a program in MicroSoft C v4.0 and I am trying to spawn to command.com to allow a user perform DOS commands (2.10). The problem is that the environment is filled with garbage and returns with environment full error if I try to alter paths etc. Here is a sample of what I am trying. spawnle (P_WAIT, "command.com", "command", NULL, environ); I have found that if you add the path of command.com before the NULL the environment is clean but all previous paths are lost and I would like to retain them. Any suggestions or reference books? Thanks Jamie BUWYATT@Brock.ean ------------------------------ Date: 26 Mar 87 13:28:22 GMT From: "stefan vogel" <K524911%CZHRZU1A.BITNET@wiscvm.wisc.edu> To: INFO-IBMPC@C.ISI.EDU Subject: Cross assembler for 6502 Does anyone know about a Cross assembler for a 6502 running on a PC? We are planning to do a course in digital electronics and microprocessors and want to assemble programs for a 6502 on an AT&T 6300. Thanks Stefan Vogel Inst. for Theor. Physics University of Zuerich CH-8001 Zuerich Switzerland K524911 at CZHRZU1A.BITNET ------------------------------ Date: Thu, 26 Mar 87 09:12:15 est From: sr16#@andrew.cmu.edu (Seth Benjamin Rothenberg) Subject: VMS/Vax as Mac PC file server Cc: sr16#@andrew.cmu.edu (Seth Benjamin Rothenberg) Has anyone used Vaxen as file servers for Macs and PC's together? I have been asked to find out all the options available (or soon-to-be) for using VMS Vaxen as file servers for IBM-PC's and Macintosh's. The network would have several clusters of Macs and PCs, each with a uVax. The uVaxen would be in different buildings, linked to each other and to a Vax mainframe with an Ethernet (probably DecNet). If anyone has hooked such a system up, I'd appreciate if you could let me know. Thanks in advance. Seth Rothenberg sr16@andrew.cmu.edu ------------------------------ Date: THU, 26 MAR 87 16:13:11 BST From: PETER%UK.AC.SALFORD.SYSC@ac.uk Subject: Slowing Autorepeat on Keyboard Dear IBM Gurus, Does anyone have any idea how to slow down or stop the auto repeat on an IBM keyboard from masm. Ideally I would like to be able to lock the keyboard and stop the keyboard buffer being filled at will. If anyone has any ideas I would be obliged if they could send me a copy of their reply direct, as well as sending it to info-ibm-pc. In the hope that someone may be able to help, Regards Peter Atkins, Computer Centre, University of Salford, England PETER%SYSC.SALFORD.AC.UK @ CS.UCL.AC.UK ------------------------------ Date: Thu 26 Mar 87 18:07:01-PST From: QUILES@USC-ECLB.ARPA Subject: dDBASE III Frequency Analysis Wanted I wonder if I can get some help on a dBASE III problem. The problem is this. I need to fine a way to do some frequency analysis with data in dBASE III. If your familiar with SAS what I need is something like a PROC FREQ. I'm really looking for a program already written to do this for me in dBASE III. If I'm unable to find such a program, then I'll have no other choice but to send the data to Mini-Tab or PC-SAS. Can anyone help with the whereabouts of such a program. Ruben Quiles ------------------------------ Date: Thu, 26 Mar 1987 17:52 PST From: JAJZ801%CALSTATE.BITNET@wiscvm.wisc.edu Subject: Microport Unix vs. Xenix I'm preparing to embark on Unix for my 286 machine and am somewhat confused by the variety of implementations. Can somebody familiar with Xenix and the Microport System V/AT compare the merits of one choice over the other (or any other flavors of PC-based Unix. I'm also confused by the seemingly large numbers of Xenix implementations: I thought it was a Microsoft product. Jeff Sicherman JAJZ801@CALSTATE.BITNET ------------------------------ Subject: Graphics Software Request Date: Thu, 19 Mar 87 18:13:03 EST From: 20s1-s4@braggvax.arpa The 20th Engineer Brigade is looking for two types of Graphics software; a business-oriented package to generate high-quality visual aids for our Command briefings, and a Computer Aided Design & Drafting package. Due to high personnel turnover, we would be most interested in those packages that have a tutorial. We are currently operating WANG PCs/APCs, and some IBM clones (Zenith). Available RAM is from 640K to 1.1 M; minimum Winchester capacity is 10 M. Any information would be greatly appreciated. Thanks in advance, Dave Wegener Brigade Automation Management Officer & All-around Nice Guy "Sure I'm qualified-- I had FORTRAN back in '78 !!" ------------------------------ Date: Thu 19 Mar 87 19:32:29-CST From: Michael Sheridan <SHERIDAN@STL-HOST1.ARPA> Subject: Integration of PC and VAX I am in the process of evaluating hardware/software that will enable PC users to "painlessly" share files that will be located on drives attached to a DEC VAX. Does anyone out there have any experience with any of the following: 1. XYPLEX'S PC XPANDER 2. DEC'S DECNET DOS 3. DATABILITY SOFTWARE'S RAF (REMOTE ACCESS FACILITY) All replies will be appreciated, and if there is sufficient interest, I'll post a summary. Thanks in advance, Michael Sheridan Sheridan@STL-HOST1.ARPA ------------------------------ Date: Thu, 19 Mar 87 21:53:53 est From: Mike Ciaraldi <ciaraldi@rochester.arpa> Subject: Want to Kill Keyboard Repeat I am looking for a (preferably free) program that will let me turn off automatic keyboard repeat on a standard IBM PC, or at least slow it down enough that you would have several seconds to take your fingers off the key. I tried KBFIX22, which is supposed to let you set both the delay before starting repeat, and the repeat rate, but these can only be increased to a period that is still less than a second. Five seconds or more is more what I am looking for, if not infinity. Thanks. Mike Ciaraldi ARPA: ciaraldi@rochester.edu uucp: seismo!rochester!ciaraldi ------------------------------ Date: 20 Mar 1987 08:39-CST Subject: Computer Products United Micros From: SAC.55SRW-LGS@E.ISI.EDU I'd like to hear from anyone who has a PC from Computer Products United of Pasadena, California. I've seen their ads in PC magazine. Two friends have bought from this company. In both cases, they had to return the motherboard for overheating problems. One guy got a replacement (new or repaired, I'm not sure)> The other guy just sent his in. The guy who got his back has been using it for a month, now, no more problems. Reply to SAC.55SRW-LGS@E.ISI.EDU Thanx in advance, Frank Starr ------------------------------ From: Jeff%acorn@oak.lcs.mit.edu Date: Fri, 20 Mar 1987 11:29 edt Subject: Postscript Filters I'm looking for a filter that will accept straight text input and output Postscript-embedded text for printing on Postscript printers. It would be nice if it could accept ^L's and such, even nicer if it had command line options for page length and width and would rotate (landscape) text. Preferably PD, preferably source in most any language. -jeff Jeff Collier jeff%acorn@oak.lcs.mit.edu Gold Hill Computers 617.492.2071 163 Harvard St Cambridge, MA 02139 ------------------------------ Date: Fri, 20 Mar 87 15:05 EDT From: <MGRANOFF%CLARKU.BITNET@wiscvm.wisc.edu> Subject: Finding Current Application Name I am designing a memory resident application under DOS 2.11 on an AT&T 6300. The application will be for use with another commercial product, but will only be of use if the commercial application is running (i.e. my memory resident application will be of no use at the DOS prompt). What I would like to be able to do is determine if a program is running or if the user it sitting at the DOS prompt. If *that* is possible, then is it possible to determine the NAME of the application that is running? I will be writing my application in Turbo Pascal v3.01. Any help on this will be appreciated; any code, etc. would be better in Turbo Pascal, but I think I can probably hack (I hate that expression...) anything... Thanks in advance. Mark H. Granoff MGRANOFF@CLARKU.BITNET [The file INDOS.TXT has a discussion of some of the issues involved in such a memory resident program. I am not aware that the name of the application is available anywhere. -wab] ------------------------------ Date: Fri, 20 Mar 87 11:41:01 PST From: iwamoto@cod.nosc.mil (Harvey A. Iwamoto) Subject: FileCard 20 on AT Has anyone been able to get the Exzel-Dysan Filecard 20 working with the IBM PC AT as a second drive? I have been in contact with Exzel and what they mean as 'AT compatible' is that the Filecard is installed as the C drive and the only hard drive on an AT. Who would order an AT without a hard drive? Thanks for any hints or pointers. Harvey Iwamoto [Sounds strange to me as well. The AT comes with a hard disk controller. Does the Filecard add a second controller or work off the AT controller? -wab] ------------------------------ End of Info-IBMPC Digest ************************ -------