Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Path: utzoo!utgpu!water!watnot!watmath!clyde!cbatt!ucbvax!C.ISI.EDU!Info-IBMPC From: Info-IBMPC@C.ISI.EDU.UUCP Newsgroups: mod.computers.ibm-pc Subject: Info-IBMPC Digest V6 #20 Message-ID: <8704010435.AA04069@ucbvax.Berkeley.EDU> Date: Tue, 31-Mar-87 19:05:16 EST Article-I.D.: ucbvax.8704010435.AA04069 Posted: Tue Mar 31 19:05:16 1987 Date-Received: Fri, 3-Apr-87 04:27:21 EST Sender: usenet@ucbvax.BERKELEY.EDU Reply-To: INFO-IBMPC@C.ISI.EDU Distribution: world Organization: The ARPA Internet Lines: 994 Approved: info-ibmpc@b.isi.edu Info-IBMPC Digest Tuesday, 31 March 1987 Volume 6 : Issue 20 This Week's Editor: Billy Brackenridge Today's Topics: Rage about Apple's PC AppleTalk Card & Software Rave about the AST Premium 286 PostScript Embedder 6805 Crossassembler Keyboard Speed Tempest PC MS-DOS Goes to Cleveland Speller Dictionaries ANSI.SYS Keyboard Reassignment Key Reassignment with NANSI Graphics Programs for the Army DOS Buffer Management TEKTRONIX Emulator Summary Everex HD/FD Controller EVEREX EV-390 HARD FLOPPY CONTROLLER MicroEMACS TallTree JRAMdisk and German Keyboard Today's Queries: Russian Word Processors Zenith 286 Machine Screen Scroll Speed on ATT6300 Statistics Programs Accelerator Boards NEC V40 V50 ---------------------------------------------------------------------- INFO-IBMPC BBS Phone Numbers: (213)827-2635 (213)827-2515 Date: Fri, 27 Mar 87 23:43 EST From: Hess@MIT-MULTICS.ARPA Subject: Rage about Apple's PC AppleTalk Card & Software cc: info-mac@SUMEX-AIM.STANFORD.EDU (Alas, you read right -- "rage", not "rave". At the end of this flame is a request for help from somebody who makes AppleTalk cards & software.) 1) It *does* work. I can print to my LaserWriter. 2) The software user interface is mediocre. The driver base is way too big to be a TSR. 3) The documentation is either incomplete or afraid to say what needs to be said. <> First, the AppleTalk driver, ATALK.EXE does a Terminate and Stay Resident and takes up 58K! Centram's ATALK.SYS only takes 19K plus buffers. Why? Second, all the printing programs look for LW.ENV only in the current directory. I can see the words "PATH" and a search directory name in the .EXE, but neither seems to be looked for. (So, OK, I hacked the binary to look in "\LWENV" instead of "LW.ENV". Shouldn't have to, though. Of course, neither documentation nor switch-listing says anything about PATHs, or the fact that the LW.ENV has to be present.) Third, why does a printing program clear the screen, displaying status messages sprinkled around the 24 lines? And why does the "-q" (for "quiet") switch still clear the screen, and print even more messages, just not sprinkled around? Finally, the manual says the card comes set up to use COM2's IRQ 3. Can be changed to IRQ4 for COM1, or to IRQ2. Says not to use IRQ2 in an AT. Why does it say that? Maybe because of hard disk interference? Well, they ought to say so, as warning, or else include the sentence that follows logically from their current instructions: "You can't use this card in a PC/AT where you also need to use both COM1 and COM2." Of course, it seems to work just fine on IRQ2. At least my AT hard disk hasn't slowed down or died just yet... Of course, Centram's card comes set up for IRQ2 and says nothing bad about installing in an AT, nor does their tech support see any reason why not. (I don't know who's worrying/not worrying appropriately...) <> I'm sending this message to both Info-Mac and Info-IBMPC because the card may be of some concern to each group, and in hopes of catching the eye of someone from Centram, Apple, and Tangent. Or anybody from ThInK who's had to use all these cards working on InBox and might have hints. Regardless of my flaming, I do need some help: * Will Apple improve its software? Is there a reason it can't get small? * If not, will Centram please do a version of TOPS that runs on the Apple card and sell it separately? * If not, does the latest Tangent software (which seems to be underneath the Apple logo in this case) work any better than this? * If not, should I know better than to buy new toys? :-) Brian ------------------------------ Date: Sat, 28 Mar 87 00:51 EST From: Hess@MIT-MULTICS.ARPA Subject: Rave about the AST Premium 286 I said so many bad things recently that I wanted to say something good! Well, I've not used a better machine than the AST Premium 286. It's a 10MHz no-wait AT clone box. Mine has 3Mb of RAM (private bus to get the no-wait operations) and a 70Mb hard disk. Comes with an EGA standard equipment. All the stuff in my machine is AST-labeled equipment. It has performed flawlessly. The reason I'm so excited about it is that doing Dhrystones, it's just under 6 times as fast as a PC. (For the benchmarks with the compiler we use, that makes it over 2 times as fast as a Sperry IT running in AT 6Mhz compatibility mode, and just under twice as fast as an Orchid TinyTurbo 286 card in a PC (all the memory references must have been in the cache RAM...).) Subjectively, it flies! Brian ------------------------------ Date: Sat, 28 Mar 87 12:44 EST From: Hess@MIT-MULTICS.ARPA Subject: PostScript Embedder To: jeff%acorn@LIVE-OAK.LCS.MIT.EDU Well, if you don't find anything public domain (though I think Info-PostScript had something like this) you have a few options: (1) Walk over to jlh%acorn@live-oak and ask to see how FinalWord II works with the "-plain" option for printing source files, and ask him about the "LaserWriterLS" printer type for rotated text. (2) Regardless of what I said about Tangent Technologies' software for actually driving the AppleTalk card and sending Laser printer output, they certainly do seem to have the conversion process licked, for ASCII as well as a host of common WP format files. But these two are commercial solutions. If you want free, the best thing is to ask Info-PostScript or Laser-Lovers for a PostScript program (rather than an IBM PC program) that takes ASCII as input and does all the work in the printer. Brian ------------------------------ Date: Sat, 28 Mar 87 10:39:50 CST From: galvin@sally.utexas.edu (Peter Galvin) Subject: 6805 Crossassembler I just placed the file UASM.ARC on simtel20.arpa in PD: This file contains the C sources and executables for 8051, 6805, and Z8 crossassemblers. The package is shareware. I'm working on a 8096 version of the package. Since you're on the bitnet, you'll need to request the file via mail to the simtel file server (directions previously posted by Keith Peterson). --Pete ------------------------------ Date: Sat, 28 Mar 87 10:46:58 CST From: galvin@sally.utexas.edu (Peter Galvin) To: PETER%UK.AC.SALFORD.SYSC@ac.uk Subject: Keyboard Speed I don't know about slowing down the keyboard, but if you're just trying to avoid filling the minimal keyboard buffer there are any number of utilities to stretch the buffer to up to 160 characters. The file SIMTEL20.ARPA::PD:BUF160.ASM comes to mind. Sorry if this isn't what you're looking for. --Pete ------------------------------ Date: Sat, 28 Mar 87 09:21:27-1000 From: Kent K. Kuriyama To: dowen@bbng.ARPA Subject: Tempest PC Zenith sells the government a tempested version of their Z-150's for about $3,500. If you can wait, I'm told that Zenith is coming out with a tempested AT. Zenith's number is: 1-800-582-0030. Kent Kuriyama Naval Ocean Systems Center Hawaii Laboratory ------------------------------ Date: Sat, 28-Mar-87 17:49:16 PST From: ghsvax!pelican!pete@csvax.caltech.edu (Pete Carah) Subject: MS-DOS Goes to Cleveland To: jfjr@mitre-bedford.arpa Your C problem sounds like one I had several times when linking C programs with assembly programs. There are two main cautions and a couple more minor ones: 1. Make sure that the stack is left in the condition that the compiled code expects. This means (for C only) that you PUSH BP MOV BP,SP SUB SP,tempsize ; if you use stack temps . . . MOV SP,BP ; This can be an add sp,tempsize, and isn't ; always needed. POP BP RET ; THIS RET HAS NO OPERAND POPS Popping arguments off the stack is done by the caller. For a model of this, see a disassembly of a short compiled subroutine using the memory model that you use in your actual program. 2. Make sure that all offset, dw, or dd references to default data segment addresses have DGROUP: in front of the symbol. Failure to do this will result in incorrect offset fixups by the linker. Note that the lea instruction doesn't need this as long as your assume is correct. This one will normally not cause a lockup unless you have indirect branches through data segment variables. 3. This one bit me with no warning (using 3.0 C and its linker) - I don't know if it is still there but it may well be: There needs to be an assume ds:DGROUP before the _DATA segment. Failure to do this leads to mismatch between C and assembly references to symbols that are defined in the assembly module. This will normally not cause a lock-up but if there are any references to function addresses passed in extern variables, it could. This one appeared to be a linker bug and may be fixed with 4.0. It doesn't hurt to include this assume, however. 4. One of the cuter lock-ups that I had resulted from the following: printf("%s %ld\n", "string"); (leaving the long argument out). Apparently printf used the variable location in the stack as a temporary, thus wiping out the callers BP and IP. I don't know if the 4.0 printf still does this, but it is common practice in C to use an argument as a temporary. These are a few pieces of Microsoft C folklore that I have learned (the latter two the hard way). It is useful to have a real reset switch anytime you are debugging interrupt routines. They have a way of needing it. The instructions for this switch have been described in several back issues (grounding the rear-most power supply pin via a spring-return switch works for real IBMs and most clones. The correct wire is labeled power ready in the schematic). I use a center-off switch with the other direction connected to NMI. -- Pete ------------------------------ Date: Sun, 29 Mar 87 14:28:02 EST From: Brent W Baccala To: jperry@sri-unix.arpa Subject: Speller Dictionaries Brown Bag Word Processor/Mail Merge/Speller (called Word Processor with MailMerge) has a 50,000 word vocabulary (expandable). Its $130, runs on IBM PCs and most compatibles (or so I'm told). My mother (who does a lot more word processing than I) is very pleased with it. It's also impressed me on the relatively few times I've had occasion to work with it. I don't work for Brown Bag, and here is their address: Brown Bag Software 2105 South Bascom Avenue, Suite 164 Campbell, CA 95008 (408) 559-4545 - BRENT W. BACCALA - Computer Aided Design/Interactive Graphics U.S. Naval Academy Annapolis, MD [INFO-IBMPC is spell checked by Micro Spell. Aside from a few CPMisms it works fine for me. Unfortunately the copyright notice doesn't give the address other than Bob Lucas Trigram Systems. I think they are in Pittsburgh or Chicago. Last month's Personal Computing magazine has a very good article on Spell checkers. -wab] ------------------------------ Date: Mon, 30 Mar 87 08:03:33 CST From: mlw@ncsc.ARPA (Williams) Subject: ANSI.SYS Keyboard Reassignment I have to disagree about ANSI.SYS having nothing to do with keyboard reassignment. In fact, here's some stuff from the IBM Technical Reference for DOS 3.2... Keyboard Key Reassignment The following table contains the control sequences you can use to redefine the meaning of keyboard keys. ESC[#;#;...#p or ESC{"string";#;#;"string";#p or any other combination of strings and decimal numbers Function: The first ASCII code in the control sequence defines which code is being mapped. The remaining numbers define the sequence of ASCII codes generated when this key is intercepted. However, if the first code in the sequence is 0 (NULL) the first and second code make up an extended ASCII redefinition (see Chapter 6 for a list of extended ASCII codes). [sorry about the non-tabular presentation...-mlw] Here are some examples: To execute these examples, you can either: -Create a file that contains the following statements and then use the TYPE command to display the file that contains the statement. -Execute the command at the DOS prompt. 1. Reassign the Q and q key to the A and a (and the other way as well): Creating a File: ESC [65;81p ESC [97;113p ESC [81;65p ESC [113;97p This material was taken from pages 3-17 and 3-18 of the IBM DOS Technical Reference Manual. More examples are present there. Mark L. Williams (mlw@ncsc.arpa) ------------------------------ Date: Tue, 31 Mar 87 10:27:12 CST From: mlw@ncsc.ARPA (Williams) Subject: Key Reassignment with NANSI Billy: Apparently the keyboard reassignment IS supported by nansi. Here's an excerpt from our nansi.doc file, as downloaded some time ago from simtel20... IBMKKR Keyboard Key Reass. ESC["string"p The first char of the string gives the key to redefine; the rest of the string is the key's new value. To specify unprintable chars, give the ASCII value of the char outside of quotes, as a normal parameter. IBM function keys are two byte strings; see the IBM Basic manual. For instance, ESC[0;";dir a:";13;p redefines function key 1 to have the value "dir a:" followed by the ENTER key. | If no parameters given, all keys are reset to their default values. The only problem with answering questions on the system is not knowing how many answers some questions will prompt. It's too bad to be so busy that one can't respond, yet drafting responses that duplicate several others is obviously unproductive. Ah, well. Some day we'll all know everything, right? Later... Mark (mlw@ncsc.arpa) [The editor eats crow. I was wrong. I guess ANSI and NANSI do allow key reassignment. Please keep those cards and letters coming, better too many responses than none at all. In this case my original comment that ANSI didn't do keyboard reassignment went unchallenged for several weeks. Glad to see there is some reality test out there. NANSI.CAT (source code for an extended ANSI.SYS can be downloaded from either of our lending libraries. -wab] ------------------------------ Date: Mon, 30 Mar 87 08:11:54 CST From: mlw@ncsc.ARPA (Williams) Subject: Graphics Programs for the Army Re: presentation and CAD graphics for the Army...If you utilize the existing microcomputer contract with Zenith Data Systems, you can acquire Graftalk version 2.0 for $81 and Cadkey V2.0 for $280. Note that CadKey lists for about $2600. No mistake on the decimal, there. Hardware delivery into the Navy has been rather poor (although within the contract requirements), but software should be more quickly obtained. Zenith has been upgrading its capabilities, too, and even hardware is supposed to be delivered in a more timely manner. If you need the contract number or CLIN info, let me know. Mark L. Williams (mlw@ncsc.arpa) ------------------------------ Date: 30 Mar 87 11:48 EST From: DIXON%CSBVAX.decnet@ge-crd.arpa Subject: DOS Buffer Management I cannot speak authoritatively, but I have spent a reasonable amount of time disassembling DOS (v3.1) for various reasons. While I would not go so far as to call DOS caching brain-damaged, it is hardly intuitive. The size of cache blocks is fixed at boot. IBMBIO contains the basic device drivers (CON, $CLOCK, disk) needed to complete the boot process. After some initialization, IBMBIO calls IBMDOS, passing the listhead of the device chain in DS:SI. IBMDOS scans this device chain and performs various housekeeping functions. Among other things IBMDOS builds device control blocks for the currently existing disk drives and calculates the largest sector size. The first cache block is allocated and the cache block listhead is initialized. Much of the work IBMDOS does is to build a data structure which is passed back to IBMBIO. This data structure contains the cache block listhead, the number of block devices, the device control block listhead, and the size of the largest sector (which, I believe, fixes the size, of all cache blocks). At this time enough of DOS is in place (INT 21 handler and basic device drivers are in place. I have not disassembled IBMBIO, so I am only speculating what happens at this point. I believe when IBMBIO starts executing again, it opens and processes CONFIG.SYS. When a BUFFERS= statement is encountered, IBMBIO allocates the required number of cache blocks. to IBMBIO fixes the cache block size. Cache blocks are maintained in a singly linked list. Each block contains one sector's worth of disk data plus 1 paragraph of overhead. The overhead area is used for links, block status, and block identification. The head of this list is located at a known location. DOS searches this list sequentially, beginning at the listhead. Blocks are allocated from the front of the list and generally inserted at the tail of the list. DOS maintains file context information (file position, access, etc) in a data structure called the system file table (there are actually two SFTs' -- one for FCB access and the other for handle style access). When a read request is made for an open file, DOS uses the handle to get to the appropriate SFT entry. At this point the request is split into three separate parts: (1) If the request does not begin on a block boundary, the number of bytes to the end of the first block. (2) The number of complete blocks (3) If the request does not end on a block boundary, the number of bytes needed from the final block. If the first part of the request has a non-zero length, a cached read is attempted (i.e look in cache first -- if its not there, read it). DOS breaks the second part of the request into physically contiguous blocks and reads these blocks directly into the user's buffer. After each read completes, DOS scans the cache to see if any of the blocks just read in are also in the cache. If the blocks are cached and are dirty (i.e the incorrect information was read from disk), the contents of the cache blocks is copied to the user buffer. Any cache block which would have satisfied the request (whether dirty or not) is moved to the end of the list of cache blocks). The final part of the request (if non- zero length), is handled with a cached read(? - my memory is a little hazy here). Although there are better ways of handling caching, there is some logic to this algorithm. Block device drivers can only deal with whole blocks, so DOS must come up with buffers for any partial blocks. The buffer cache is used as a buffer pool. Most file access is sequential. If a request does not begin on a block boundary, it is likely that the block has been referenced before. Similarly, if a request does not end on a block boundary, it is likely that the block will be referenced again. Copying individual complete blocks is time consuming and DOS has no basis for suspecting that a complete block request can be satisfied from cache. (Did I hear somebody say RAM disk?) DOS does not use cache for full block reads. (Of course a check is made for dirty blocks). I hope that this explanation helps. Walt Dixon (Usual disclaimers) ------------------------------ Date: Mon, 30 Mar 87 14:28:29 +0200 (Central European Summer Time) From: XBR1YD2F%DDATHD21.BITNET@wiscvm.wisc.edu (Ralf Eberhardt) Subject: TEKTRONIX Emulator Summary Hello, Two weeks ago I requested info about available Tektronix emulators for the IBM-PC. BTW: I have no access to any databases where I could get information on that topic from our EARN-node. I think it is a good idea to summarize all the answers to the list, since there seems to have been a lot of interest. Here is a short summary of all the products about which I got answers: (Thanks again to all who where so helpful.) 1. VTEK from Scientific Endeavours Corp. Route 4, Box 79, KINGSTON, TN 37763 Phone: (615) 376-4146 U.S.A Description: Tek 4010/4014 emulator, runs on PC, XT or AT and can be configured for various peripherals, among them EGA also. It has VT100 emulation also. They charge $150 for it. The documentation is simple but not much required. You will not get true colors with this program, i. e. this is not an emulation of TEK 41xx terminals. Original mail from: Reuven Weiss (REUVEN@TAUENG) 2. TGRAF-07, which is a TEK 4107 emulation from Grafpoint Inc., San Jose, CA, USA. This works on a 256KB EGA or C&T compatibles. (no further info) Original mail from: Reuven Weiss (REUVEN@TAUENG) 3. QKKERMIT available at Queens Univ. (contact VIC@QUCDN) Description: This is an implementation of KERMIT with VT100 and TEK4010 terminal emulation, written in Turbo Pascal. Original mail from: Kevin Lowey BITNET: LOWEY@SASK (preferred) UUCP: ...!inhp4!sask!lowey 4. EM4010 by Diversified Computer Systems Description: A very good Tek4010 emulator. It's emulation is one of the fastest 4010 emulations and it also emulates VT100. Other features include XMODEM, Plotter and printer hardcopy, Logging of screens, Image capture (stored as a Tektronix instruction stream) and support for EGA and a few other graphics cards. The cost to an educational institution is $69. US in quantities of 10 or more. The address of DCS may be found in a recent copy od PC WORLD. Original mail from: Kevin Marinelli (marinell@DAL) Academic Computing Services Halifax, Nova Scotia Canada 5. PerSoft Software in Madison, WI, has a series of VT-100 emulators (their SmarTerm line) which have vt220* and Tek graphics emulation They are continually rated #1, they have file xfer, graphics, keyboard mapping, all round *good* software! Address: Persoft 465 Science Drive Madison, WI USA 53711 +1 608 273-6000 Original mail from: John Plocher (plocher@puff.wisc.edu) 6. a) For the Macintosh: VersaTerm and VersaTerm Pro from Abelbeck Software. The former is VT100/Tek 4014 emulator and the latter is a VT100/Tek 4105 emulator. "Major" drawbacks: no color, no direct support for plotters. They work, have a remarkable number of useful bells and whistles, and take full advantage of the Macintosh interface. I generally prefer PCs to Macs, but these emulators are generally much better than anything I've found for PCs. Mac240 from White Pine Software. A VT240 emulator. I have not yet tested it. b) For PCs and compatibles: EM4010 from Diversified Computer Systems. A VT100/Tek4010 emulator. I have not yet tested this, but NCAR (National Center for Atmospheric Research) likes it. PC-PLOT from MicroPlot Systems. A VT100/Tek4010 emulator. Not much flash (no pan, no zoom, no integrated file-transfer protocol), but it works well. It works with many PC-compatibles, monitors, printers, and plotters; the manual distinguishes between machines that definitely do, definitely do not, and may work (MicroPlot is not sure). PC-PLOT is also a partial Tek4027 emulator; I've set up DISSPLA and TELL-A-GRAF to avoid hardware fill, and they work well with PC-PLOT in 4027 mode. I suspect SASGRAPH can be set up similarly. SmarTerm 240 from Persoft. A DEC VT240 emulator. I don't have software that can give this a thorough test, but it passed my simple tests. Supports several printers (including HP LaserJet) and HP 74xx plotters. TGRAF-05 and TGRAF-07 from Grafpoint. The former is a Tek 4105 emulator and the latter is a Tek 4107 emulator. I had a copy of TGRAF-07 for a limited time and did not have software to really test it, but it is full of flash. These are expensive, but good. ZSTEMpc-VT100 and ZSTEMpc-4014 from KEA Systems, Ltd. The former is a VT100 emulator and the latter is a Tek4014 emulator. Get just the former if you want only VT100 emulation; get both if you want 4014 emulation (when used together, they work essentially as one program with overlays). KEA claims they have the best VT100 emulator available; I couldn't do a thorough test, but it does seem better than most people will need. The 4014 emulation is good, and has pan and zoom. Original mail from: Robert Zaret (ZARET@MITVMA) 7. VTERM/4010 is a commercial product. It can emulate a VT100, a VT52 and a Tek 4010. (Actually, I think it may be a superset of the 4010 such as a 4014 but I am not sure.) It also has a "hydrid" mode that switches between VT100 and Tektronix upon receiving escape sequences which indicate what output is received. I have used this hybrid mode for the most part. One host I use works so well that it switches VTERM at all the right times. If this does not happen for you, there is a key to press to switch modes. You can order VTERM/4010 from the authors. Their address is: Coefficient Systems Corporation 611 Broadway New York, NY 10012 USA telephone: (212) 777-6707 The cost is about $250 U.S. They also have terminal programs that do not include the 4010 portion so be sure to specify that you want it. Twice they have sent me the cheaper version. Original mail from: Tom Reingold; The Rockefeller University; 1230 York Av; NY 10021 Ralf Eberhardt, (ARPA: xbr1yd2f%ddathd21.bitnet@WISCVM.WISC.EDU) (BITNET: xbr1yd2f@ddathd21 ) Technical University Darmstadt, Computing Center, Darmstadt, West Germany ------------------------------ Date: 1987 Mar 30 19:26 EST From: Bob Babcock Subject: Everex HD/FD Controller One possible source of your problem with the floppy controller not seeing the A drive is that not all floppy drives use the same sense of the ready signal on pin 34: either high or low may mean ready. (I know it seems like this should be standard, but I have a mixture of both types of drive in my Sanyo 555, and I too pulled some hair figuring out why the new one wouldn't work.) Anyway, if this is the problem, quickly opening and closing the drive door while trying to do a DIR may make it work. Then, it may be sufficient to break the pin 34 connection to the drive and tie the controller end high or low. It worked for me, but probably depends on the particular controller. My drive even had a soldered-in jumper on pin 34, so changing the connection was easy. ------------------------------ Date: 31 Mar 87 16:13:35 GMT From: tom%unirot.UUCP@seismo.css.gov (Tom Giacchi) Subject: EVEREX EV-390 HARD FLOPPY CONTROLLER Organization: Public Access Unix, Piscataway, NJ An answer to your question is probably this: 1> Check the drive select setting on your floppy. It should be set to DS0 if you have a straight thru floppy cable (no twists or flips in cable). It should be set to DS1 if you have a twisted cable. Also make sure that the TERMINATING RESISTOR is installed. Sometimes this is a jumper labeled TM. Oh, Some drives may start the counting of drives at DS1, if yours does then DS0 would be DS1 and DS1 would be DS2. Hope this helps. ihnp4!tarpon!orca!tom ------------------------------ Date: Mon, 30 Mar 87 23:44:14 PST From: larry@Jpl-VLSI.ARPA Subject: MicroEMACS MicroEMACS is available from various sources in several versions. V30 was (I seem to recall) Dan Kegel's version of Dave Conroy's original program. It is available from SEISMO in mod.sources/Volume8 under the name MicroGNU, if memory serves. However, you may prefer to get the latest version by Dan Lawrence. V3.8b is on SIMTEL20 in PD:EMACS*38B.ARC. V3.8f can be gotten from The Programmer's Room bulletin board at (317) 742-5533. The BBS is usually available 24 hours at 300/1200 baud. If you use KERMIT be sure to set your end to 8 bits, no parity. I have 3.8f and it is nice. However, you must compile it with the large memory model and it takes up around 100 KBytes. That includes a fairly complete macro extension language, regular expression find and replace, file en- and de-cryption, automatically sensing and adjusting to the graphics adaptor. In comparing Conroy/Lawrence's microEMACS to the others on PC's, I'd say it comes in behind Epsilon and UniPress EMACS, but not far behind. I use it because I have to swap between AT's and XT's, several supermicros, VAXes under both Unix and VMS, and two different mainframes. Lawrences's uEMACS compiles and runs on all of them, so I don't go crazy when I swap from one to the other. (It also runs on Ataris and Apples, but that's not relevant to me.) Larry @ jpl-vlsi.arpa ------------------------------ Date: Tue, 31 Mar 87 16:04:00 MEZ From: UNM406%DBNRHRZ1.BITNET@wiscvm.wisc.edu Subject: TallTree JRAMdisk and German Keyboard A few weeks ago, I asked if anybody could help us with the following problem: we had installed a TallTree JRAM disk and the warm boot (which should *not* destroy the content of the disk) didn't work. Thanks to Christian Flury from Zurich, who knew the solution: We use a german keyboard, which is supported by a keyboard driver KEYBGR.COM (from IBM). This driver produces several symbols via Alt-Ctrl-Something sequences and therefore looks for such combinations. It catches Alt-Ctrl-Del and blocks the combinations which are used by JBOOT. So much for that. Thomas Miller(UNM406%DBNRHRZ1.BITNET) ------------------------------ Date: Mon, 30 Mar 87 22:56:40 PST From: porter@Juliet.Caltech.Edu (Mark A. Porter) Subject: WP in Russian? I need to do word processing using the Russian (Cyrillic?) character set. I need it to display the characters in some reasonable format on the screen, and to print to most reasonable printers (IBM Proprinter specifically). I am normally a Vaxen type of person, so please treat me as if I know nothing of the subject. Unfortunately, this program needs to run on the IBM PC/XT/AT, under PC/MS-DOS. I need to know: 1. Who makes these products? 2. What are the issues involved? 3. What hardware must I have? Thanks in advance for any help! ~Mark ~~~~~~~~~"Mustang"~~~~~~~~~~~~Mark A. Porter~~~~~~~~~~~"Untamed!"~~~~~~~ Caltech 1-59, Pasadena, Ca, 91126 (818) 792-3711 ...!cit-vax!porter PORTER@CITJULIE.BITNET, PORTER@JPL-VLSI.ARPA, PORTER@JULIET.CALTECH.EDU [Since this is the second query in as many digests I reprint the response referenced in the last digest. V5 #20 -wab] ------------------------------ Date: Tue, 11 Feb 86 01:40:21 pst From: cole@Babette.isi.edu Subject: Russian (and French) Word Processors I asked an associate (a fellow by the name of Potashnik) about word processors for Russian, and he gave me the name of one company who makes one. The company is called Economic Insights, of Rockville, MD, phone (301) 258-5235. Their product uses a Quadvue display board, generates output only for Diablo 630 and Epson LQ-1500 printers, and costs $595. That's all I know about them (I don't speak or write Russian and never heard of the company until now). On a (remotely) related subject, another friend who is a big fan of Microsoft Word is in the process of getting a French version of MW. It is apparently only available in France, however. He talked to Microsoft in Canada and they have a French-Canadian version, but it is said to be significantly different from the French version. I don't know if there are any other versions out there for other languages, although I doubt they have a Russian version. If they did you might have to order it from the USSR and pay for it in rubles, anyway. Randy Cole Citicorp-TTI ------------------------------ Date: Sun, 29 Mar 87 20:46 EST From: OPER013%UMUC.bitnet@BERKELEY.EDU Subject: Zenith 286 Machine Does anyone have experience with Zenith's 286 machine? With 20 meg, ega card, 640, and the educational discount the machine prices to $2,299. We are looking to setup 2-3 labs with AT class machines and I really don't want Vectras. That is what the ppl in house are after. On another subject, has anyone seen/heard/own a copy of procomm 2.4.3? the cursor run on w/ version 2.4.2 is a massive pain. nick ------------------------------ Date: Mon 30 Mar 87 01:41:33-PST From: Jean-Pierre Dumas Subject: Screen Scroll Speed on ATT6300 Can anybody help/give a solution/explain why the scroll speed on my Olivetti M24 (AT&T 6300) is at least 4 times slower than on a IBM PC ??? I use: M24, 10meg HD, Monochrome, 640K, I compare to the same, but made by IBM. Both are connected to X25 network at 4800 or 9600 bps. The big problem shows when I try to "capture" files on the fly, at 4800 bps no problem on the IBM, only half the file with the M24. I can't use xon/xoff control flow, so don't tell... Other way to diagnose the phenomena: do a type of a file, and compare the scrolling speed... huge difference. I need to be able to capture files a 4800 bps at least... I use kermit version 2.29b or 2.29M24. It is advertised as fast. Jean-Pierre H. Dumas (dumas@sumex-aim.stanford.edu) ------------------------------ Date: Mon, 30 Mar 87 21:34:12 CET From: Eberhard W. Lisse Subject: Statistics Programs Can someone look up in the KWIC index the article about the statistics programs adapted from unix which was offered some time ago ? I would like to contact the author but have no index here. Thanks in advance. el [Here they are: 5.077 Statistical and Graphical Libraries Query 4.046 Statistical Freeware 5.003 Statistical Package for IBM-PC Query 5.004 Statistical Package for the IBM-PC 2.090 Statistical Package micro-TSP 4.134 Statistical Package PowerStat 3.056 Statistical Package SYSTAT 3.063 Statistical Packages (2 msgs) 4.045 Statistical Programs Query 1.027 Statistical Software 1.028 Statistical Software 2.040 Statistical Software Compaq Problem 8087 MICRO-TSP 5.007 Statistical Software for IBM PC Mathematical and 5.098 Statistics Packages for PC's 2.083 Statistics Packages Wanted 4.100 Statistics Scientific Plotting and 4.016 Statistics, Graphics and Cal Software The ones from last month haven't been added yet. -wab] ------------------------------ Date: 31 Mar 1987 05:48:09 PST Subject: Accelerator Boards From: Laurence I. Press I am looking for accelerator boards (80286 based) for Leading Edge clones and model-1 (64K motherboard) PCs. Any recommendations? Any boards to avoid? I'll summarize responses if there is interest. Lar ------------------------------ Date: Tue, 31 Mar 87 11:40:31 CST From: munnari!augean.oz!tnemeth@seismo.CSS.GOV (Tom Nemeth) Subject: NEC V40 V50 Organization: Engineering Faculty, University of Adelaide, Australia I would like to obtain information on where to obtain the new NEC V40 and V50 chips, together with prices and other relevant details if possible. Please reply to me by mail. Thanks in advance, Tom Nemeth ------------------------------ End of Info-IBMPC Digest ************************ -------