Jump to content
IGNORED

Geneve History


etownandy
 Share

Recommended Posts

For those interested, here's a history of the Geneve that I've yet to finish. Most of the information was gleaned from microPENDIUM.

 

July 1986

Lou Phillips told MP that, while not taking orders, the Geneve was on schedule for it's end-of-July launch. "BASIC 3.0" and "DOS" were mentioned, as were BASIC and C compilers.
Also noted was that three cartridges were known not to work. Two in particular were mentioned. These were Statistics and Personal Record Keeping. Chess, although not mentioned, was the third cartridge with compatibility issues. Statistics and Personal Record Keeping both made calls to the BASIC ROMs of the machine, ***ALTHOUGH IT IS UNCLEAR WHY THAT MAKES THEM INCOMPATIBLE WITH THE GENEVE.
Others were later found to be incompatible, although most of the incompatibilities were due to earlier cartridges only having an option to save to cassette - something the Geneve didn't support. Additionally, many cartridges printed to "RS232" (the serial port) instead of "PIO" (the parallel port), although this was much easier to fix than the load/save device. Often, sector editing the cartridge dump and replacing any instance of "RS232" with "PIO" was sufficient. A final software incompatibility arose with software that expect to use the TMS9918A video display processor. Some software was written using VDP register values that were actually out of range. The 9918 would "wrap" the overflowed value and "loop around" to the correct value if there was only 16K of video memory. Since the 9938 in the Geneve treated these as VALID values, unexpected results would occur usually resulting in a scrambled display.
August 1986 - masthead changes; an ad from Video Electronics Marketplace announces "September Delivery"; newsbyte feature "confirmed" the September delivery with a note from Lou Phillips that the hardware was complete but software was still being worked on.
Sept 86 - first rumors of device being developed by Miller's Graphics to let a TI emulate a "PC running a PC DOS-type operating system."
Oct 86 - Announcement of HFDC
Nov 86 - Article describing demo of Geneve by Phillips at Chicago Faire. Phillips noted production was being held up due to the availability of the custom-designed gate array chip which "contains all the logic, the dynamic RAM refresh logic, the wait state logic so we can run this machine at various speeds...It also handles the memory mapping that the 9995 uses to access more than 64K." Phillips was assisted in this demo by J Peter Hoddie and Paul Charlton. Hoddie was mentioned as working on modification for TI-Writer (which would ultimately become MY-Word), adding 80-column support, a View File command, and more. Note Assembly language guru Mack McCormick was working on the 80-column patch for Microsoft Mupltiplan. Pecan Systems was announced as the developer of the ultimately never released PASCAL runtime (v4.21) which was meant to be used to run various compilers. Pike Creek Computers, makers of TI Count, were reported to be working on business software. Paul Charlton was mentioned to be rewriting his Fast-Term terminal emulator for the machine, and Insecbot of TI-Artist fame were said to be writing a "MacPaint equivalent." Clint Pulley, author of the c99 "small c" compiler for the 4A was mentioned as creating a "real" C compiler for the Geneve.
The amount of re-development was due to the nature of the Geneve itself. Although the machine could run almost all software designed for the 4A, it required the use of the "GPL Interpreter" program. Additionally, software run through GPL could not take advantage of the Geneve's unique abilities, especially its RAM and Operating System. Some software used methods of scanning the keyboard that were non-standard (Fast-Term, 4A Talk) and such software also had difficulties.
A final piece of software was seen by some as the holy grail of Geneve development - a Lotus 1-2-3 clone with direct file compatibility between the IBM PC.
As we'll see, these development efforts also depended on having a stable and complete operating system, and the lack thereof spelled doom for the young machine.
Dec 86 - Lou Phillips announces that the gate array chip has been received from Japan and that production is starting.
MG, formerly Millers Graphics, announces the end of production of the GRAM Kracker, but announce a product which "fits in with our coming IBM expansion system for the 4A." This expansion was to be announced the following month.
Jan 87 - the monthly Myarc advertisement announced the Geneve would be "on your dealer's shelf in February."
February 87 - mp masthead changes to "Covering the TI99/4A, the Myarc 9640 and compatibles". Coverage was given to the Triton Turbo, the device mentioned by MG as an IBM "upgrade" for the TI. Triton was a large marketing company that bought a huge stock of TI inventory after Texas Instruments left the market, and was a major mail-order distributor. What the upgrade actually included was a "bridge box" that connected to the 4A's side expansion port. An IBM-style case housing the actual clone plugged in to the bridge box, as did a monitor. This basically allowed the TI's keyboard and monitor to be shared between the 4A and the clone. The advantage of such an arrangement was dubious (the 4A's keyboard was notorious for its strange arrangement and limited number of keys...common symbols like quotation marks required pressing a FCTN key then a letter key). ***MORE TURBO XT INFO***
Mar 87 - Comments column, "The Myarc 9640 computer is finally out." Although not received in time to do a review, a basic description was provided.
The usual Myarc ad announced "Available on your dealer's shelf now!"
Apr 87 - Geneve on the cover and a review starting with "The Geneve 9640 is here! Finally. And it works." Although the initial impression was very positive, several things did not yet work in this pre-production model. For starters, the word processor did not have functional search-and-replace. Extended BASIC II was included, but the machine was expected to come with XBIII.
Most glaring, however, was the lack of a finished DOS. The computer booted to a "DSK1." prompt instead of an "A:" prompt as described in the computer's manual. From that prompt, users could issue a command to slow the machine for TI software, run the GPL Interpreter used to load cartridge dumps, or start XBII.
May 87 - Part II of the review described an improved word processor (now at v2). One complaint was the the machine didn't include a boot ROM, and therefore the Operating system had to be loaded off disk. This was foreign to 99ers as the TI did not require the loading of an operating system, but was a common practice on the Amiga, Atari ST, and Macintosh - the computers most commonly referred to as Geneve "competition."
The BASIC disk was labeled "Advanced BASIC" but the title screen clearly showed this was Myarc XB II v2.11. The Personal Record Keeping cartridge dump was reported as "working fine."
This follow-up was very brief, with author (and micropendium editor) John Koloen noting, "I don't think I ought to carry this installment much further. The viability and utility of the Geneve depends on the Myarc Disk Operating System..."
July 87 - "Comments" announced that MDOS v0.8 was available for download from places like CompuServe. It was basically more a front-end for the GPL interpreter than an actual operating system.
Aug 87 - "Comments" notes that things are "getting better all the time" with improvements to MDOS and MY-Word. Also mentioned was that Myarc "has ported a 1-2-3 clone from a PC" although this never materialized.
The HFDC and MY-Art were also both mentioned.
"User Notes" featured a program to set the clock - necessary since there was no direct way to do this yet from MDOS.
Sept 87 - Geneve article debuts. Mike Dodd starts off with a description of the software bundled with the Geneve, although falling short of an actual "progress report." Reasons behind various software incompatibilities is also mentioned and numerous sector-edit patches were offered to fix these problems.
Disk Only Software announced the availability of Jumpboot. The program contained the core OS file SYSTEM/SYS on a disk that was modified to "take advantage of the advanced multiple-sector read routines of the Geneve." The 30-second load time of MDOS was reduced to 4-7 seconds.
October 87 - A list of incompatible cartridges was published and included Plato (Control Data, published by TI), Q*Bert (Parker Bros), Pole Position (Atarisoft), Buck Rogers (Sega, published by TI), Ms. Pac-Man (Atarisoft), Slymoids (TI), Early Reading (Scott Foresman, published by TI), Moonsweeper (Imagic, published by TI), Jungle Hunt (Atarisoft), TI Chess (TI).
A new partner in Myarc was announced. Jack Riley was hired by the firm on 9/1.
Mike Dodd's article focused on the two modes of operation of the Geneve - native MDOS mode and 99/4A or GPL mode. Also featured was a program to control the Myarc mouse from TI Extended BASIC (running in GPL mode).
An article announced the release of v.99 of MDOS. This was the first widely distributed version that supported all of the commands in the computer's documentation. Unlike previous versions that were posted to various bulletin boards, v.99 was to be mailed to registered users. The mailing was expected to include Advanced BASIC and Pascal Run Time 4.21. Also, Myarc Disk Manager III was being updated to allow the formatting of disks on the Geneve (a feature later brought to MDOS itself). Disk Manager IV was to be released that month and several packages that ultimately were never released were announced, including:
My-Basic - a BASIC compiler
My-Numbers, a Lotus 1-2-3 spreadsheet clone
My-Data, a Dbase III clone
a C compiler
My-Word Pro - a graphics-based word processor with mouse support
A windowing package, multi-tasking ability, and an EPROM containing the whole of MDOS were also announced as "things to look for in 1988."
A review by Mack McCormick of the Mechatronics 80-column card gushed "...if you already have the RAVE keyboard and a GRAM card or GRAM Kracker you have the equivalent of the new Geneve computer." Of course, this statement didn't take into consideration the faster CPU and enhanced RAM of the Geneve, nor the ability to easily perform file and disk management tasks from a proper Operating System, to say nothing of mouse support. In fact, a correction was made in the Comments section of the November 87 issue at the request of Myarc's Jack Riley, making it clear that the Mechatronics card would not allow a TI to run Geneve software, and that the Geneve was far faster.
Nov 87 - Rave 99 introduces a speech synthesizer adapter card for $49.95, ideal for moving the speech synth in the PE Box for use with a Geneve.
Mike Dodd's article included patches for TI Artist and Video Chess, and mentioned that some versions of Terminal Emulator II did not work. Additionally, a source code listing was provided that, when compiled, let TI Artist use the Myarc Mouse for input. Also provided was code to allow access of the Geneve clock from TI Extended BASIC.
Dec 87 - The “final” version of MDOS, v1.0, shipped. 6 minor changes were outlined in a documentation addendum. Production began on the HFDC, and versions of MDOS and Disk Manager supporting it were suffixed with an "H" on their version numbers. The Pascal runtime was mentioned as being delayed until early 1988. The Multiplan patches were also completed, and mainly offered 80 columns, faster performance, and larger (up to 41K) spreadsheets.
Millers Communications (not Millers Graphics) announced "videoflex" (text and graphics overlay) and frame grabber cards for the Geneve although the prices ($700-1,000 for the videoflex and $900-1500 for the frame grabber) were quite high. The Videoflex card was demonstrated at the Texas TI Faire in April 1988 at which time its estimated retail price was $7000-10,000.
The monthly Geneve column compared MDOS to an expanded 4A (keyboard, video card, GRAM device) in response to Mack McCormick's review of the Mechatronics 80 column card as well as an Extended BASIC disk catalog program to allow reading of the date/time stamp. Additionally, a new Q&A segment of the monthly article was started.
The My-Art/Myarc Mouse bundle was reviewed with all "A"s.
Jan 88 - Several articles this month were dedicated to the Geneve.
Miller Communications designed a $329.95 box to replace the TI PEB to be sold out of the Queen Anne Computer Shoppe in Seattle. It accommodated 8 cards including the Geneve as well as 2 hard drives and three floppy drives. A special IBM clone was also available, and the box had a switch to enable to the user to switch between computers. The 99/4A motherboard could also be moved inside the box, although naturally this required an external keyboard.
"Getting Started with MDOS" by Walt Howe dealt with the use of MDOS. Remember, the use of a proper operating system was a foreign concept to most 99'ers. The article notes that the Geneve looks for the core MDOS SYSTEM/SYS file on a Horizon RAM disk, Harddrive, or the first floppy drive. Changing current drives, getting a DIRectory, using the ASSIGN command to map letters to devices, FORMAT, DISKCOPY, and COPY commands were covered. Minor differences between the current version (1.0) and manual were also discussed.
Mike Dodd's monthly column focused on using the Geneve with a Horizon RAMdisk. Oddly, the Horizon card was better-suited to use with the 9640 than Myarc's own RAMdisk. Using MG Explorer (a debugger) and compatibility issues with terminal emulator programs was discussed, as was a "Other News" section with various items.
According to Myarc's Jack Riley, 1988 was to see the release of GPL 1.0, MDOS 1.1, Advanced BASIC, and the Pascal Runtime. The second quarter was to see the release of GEME (pronounced "Jimmy"), the Graphics Enhanced Multi-Tasking Environment, a "shell" and program launcher for MDOS as well as a relational database manager (although this is no longer being cited as a dBase clone). "Also contracted for development" was a CAD program, flight simulator, and MIDI music package. It is unclear what happened to these contractual projects.
***TALK ABOUT YOUNG PROGRAMMERS HERE***
Feb 88 - According to Bill Moseid of DataBioTics, special versions of 4A Talk and Disk Master I were being developed to take advantage of the Geneve.
Mike Dodd's column focused on programming tools, especially assembly language. Although a few independent programmers were working on shareware offerings, Myarc was developing the Programmers Development Package to include a linker, assembler, and debugger. It is interesting that Mike noted, "What software is in realistic development? The biggest two pieces of software currently under development are from Myarc: Advanced BASIC and the Pascal Interpreter." The interesting thing is that of the two packages in "realistic development" only one, Advanced BASIC, was ever finished. A general discussion of MDOS was followed by a description of various XOPs...assembly calls that "hooked" into MDOS for things like disk access.
A "Programming in MDOS" column by Walt Howe dealt mainly with the customization of MDOS via the AUTOEXEC batch file as well as with creating batch file menus.
Myarc announced a hardware-based software protection system that involved a card placed in the expansion box. Software was to be sold with credit-card sized cartridges containing a "read once" code. The cartridge would be placed in the PE box card and its code stored in the card. This was in direct response to rampant piracy of My-Art.
Also, a new Fujitsu 101-key keyboard became the standard keyboard.
A bug existed where the joystick ports were reversed. This eventually got fixed.
March 88 - Mike Dodd's regular column dealt with various problems Geneve owners wrote in about. Most useful was a list of the most recent versions to-date of Myarc software. This was:
MDOS 1.01
GPL 0.99
My Word 1.1
Disk Manager III v2.1
Multiplan 1.0
Problems with phantom duplicates of a file and comparing disks via the DISKCOPY command were addressed.
The Walt Howe article described how to create a phone autodialer out of MDOS.
An article from Bud Mills - primary reseller of the Horizon RAMdisk, provided numerous tips on using the card with a Geneve.
A couple of articles described the Myarc HFDC.
Apr 88 - Walt Howe's article focused on various strategies for organizing a hard drive while Mike Dodd's article focused on additional Questions and Answers.
May 88 - two years after the machine's announcement, the Comments feature bemoaned the lack of a finalized OS and incompatibility (later resolved) between the Geneve and the HFDC. The lack of a finalized OS was noted as being a primary threat to numerous announced pieces of software, including GEME, the Programmer's Development Package, Advanced BASIC, and more. The feature noted that these other programmers can't be paid because they can't finish their work because MDOS isn't finished.
Mike Dodd's article dealt with the sound chip (the same SN 76496 used in the 4A), cartridge memory/address space in GPL mode, and setting colors on the 9938 video chip.
The Videoflex system for the Geneve was demonstrated by Miller Communications. This was housed in a 19" rack-mount chasis complete with 80mb hard drive. It directly connected to the 9938 graphics chip of the Geneve.
June 88 - For months, MP had been asking readers to send in a list of cartridges incompatible with the Geneve they'd most like to see fixed. The list was:
Plato
Q*Bert (runs, but screen is black)
Logo II (the fix published in MICROpendium doesn't allow the program to scan a disk to load a file)
Ms. Pacman (joystick doesn't work)
Dragonslayer Spell Check (locks up prior to exiting to My-Word)
Moon Patrol (no control)
Jungle Hunt (fire button doesn't work)
Personal Record Keeping
Disk Manager II (valued for its comprehensive disk test)
Bigfoot
Dig Dug
Pole Position (fire button doesn't work)
Frogger
Donkey Kong
Computer War
Submarine Commander
River Rescue
Fathom
Other programs that readers mentioned include: Data Base Manager by Navarone, Rapid Copy, Video Graphs, Meteor Multiplication, Early Reading, Moon Sweeper, Slymoids (no graphics or joystick), Alpiner, Popeye, TEII (speech access), Statistics, Tax Investment Record Keeping, Certificate Maker 99, Moon Mine, Early Logo Learning Fun, Honey Hunt, Buck Rogers, Munchmobile, Music Maker, Jawbreak, Super Demon Attack (speech), Congo Bongo (locks up on level 2), Microsurgeon (joystick doesn't function), Henhouse (no joystick), Space Bandits (no control), and Star Runner.
It should be noted that Bigfoot, Space Bandits, and Honey Hunt are MBX cartridges from Milton Bradley, which had been previously known to be "undumpable." Programs with speech required a speech board such as that from Rave99 and GPL set to slowest (normal TI) mode but likely worked in other respects. Programs with "joystick" or "control" problems may have been the result of a known hardware bug in the Geneve - the TI standard for joysticks allowed two controllers on a single port. Thus, it was common for 99ers to have two joysticks as the TI "wired remote controllers" were a pair of sticks connected to a single plug. However, the Geneve "transposed" the joystick assignments, thus stick 1 became stick 2 and vice versa. Although some games worked with either joystick, many only worked with "joystick 1" (which would be joystick 2 on the Geneve).
Mike Dodd's article was quite ***SPECIAL PURPOSE* in that it allowed "coding and decoding of files"
July 88 - The monthly Geneve column continues to grow more specific and less general-purpose. Ironically, this month included a more detailed explanation of the "mysterious" coding/decoding program of last month.
***TIDBIT July 88 Ken Williams, president of Sierra On-Line Inc. wrote Stephen Shaw to say that, since Sierra is no longer in the 99 market, they did not "foresee circumstances under which we would enforce our TI99/4A copyrights."
August 88 - MDOS 1.06, GPL 1.02 MDM V v1.21 all released. A follow up to last month's "incompatible program" list noted that many, including Plato, Logo II, PRK, DM I, II, III, DK, Sub Comm, and River Rescue DO work. Myarc noted that they could not modify Q*Bert, Ms. Pacman or software on disk because of copyright issues. It is unclear why they could not patch these programs but could patch TI programs, when at the time, TI was still enforcing its copyright.
The first major complaints of lack of support were starting to come in. Many readers were complaining that the New Jersey office (original/main office) almost never answered the phone or that it was busy for extended periods. Mail was forwarded once a week from NJ to the Alabama office headed by Jack Riley.
One of the problems is that this was a very new system unlike anything 99ers had used before, and several users could not understand the documentation. This was perhaps reinforced by the fact that early on, the operation of the computer did not resemble the documentation at all.
The primary problem, though, was lack of resources. Myarc was a very small company, and the Geneve sold enough units to make supporting the machine an issue. Unfortunately, it would seem the price did not allow Myarc to budget for adequate support infrastructure and staff.
Myarc announced a $300 512K card for use with the Geneve that would take the computer to over a megabyte of memory. Existing 512K cards which were previously used as RAMdisks on the 4A could be returned and modified for $50 although the modification left the cards incompatible with the 4A. Instead of acting as a RAMdisk, the cards would add to CPU memory.
Advanced BASIC and Pascal Runtime were to be released in Sept, while GEME was given an October release date and $175 price tag. It would support a 512x424 resolution with the extra 512K memory or 512x212 without.
My-Pro-Art was mentioned as supporting multiple file formats, fonts, and color printing and also required the extra RAM.
MDOS 1.06 had been completed, and supported the Print Screen function and booted to 80 column mode instead of 40. Limited hard drive support was included, although full support was going to be found in v1.10H.
Sept 88
GEME was the featured item in COMMENTS and MDOS, GPL, ABASIC, and Pascal were rumored to be "finished" by October.
An article on the 9938 video chip.
MDOS 1.10 allowed use of the RAMdisk between MDOS and GPL.
Oct 88
Mike Dodd's column contained a program to redefine the MDOS colors.
Interestingly, despite GEME and My-Pro-Art being described last month as needing additional CPU memory to run high-res modes, Mike states "additional CPU memory will not increase the screen mode capability...as the 9938 will use only VDP memory." He also notes that J Peter Hoddie's MacFlix program displays at 512x424x16, higher than Macs of this era.
Starting this month was the "Myarc Q&A" column, in part a public relations move by Myarc as well as a shot at mass support of their machine - a sort of monthly FAQ.
The Q&A suggested MDOS 1.14, GPL 1.04 MyWord 1.21 Advanced BASIC and Pascal Runtime would ship in November to registered users. 1.14H would also be released, providing hard disk support. It was reiterated that he Pascal Runtime would provide access to a large library of USCD Pascal programs. Mention was made again of an EPROM based MDOS/GPL combo, eliminating the need to load either from disk and also eliminating the boot screen swan picture. "The 16K that the Swan takes up will be used to develop...special functions." Mention was also made of the importance of using compatible versions of MDOS and GPL. For example, GPL 0.99 with MDOS 1.06 or 1.08 would cause irregular operation.
Nov 88
The monthly Geneve column featured a review of T&J Software's Geneve version of DISkASSEMBLER and a debugger. Genial Computerweare's release of Paul Charlton's Picture Transfer was announced. The program could convert GIF, MY-Art, RLE, TI-Artist, and GRAPHX formats to GIF and MY-Art. A piece on computer viruses followed. At this year's Chicago TI Faire, Myarc's Jack Riley presented Paul Charlton with a plaque for his programming efforts. A patch was published for Q*Bert.
Dec 88
MDOS 1.14 released. Advanced BASIC and Pascal Runtime are "nearly finished."
A warning on mixing MDM5 files from one version to another was printed. Apparently, upgrades were sometimes posted to bulletin board systems with just the files that were changed. Users were to then replace the old versions of those files with the new ones, while leaving unchanged files alone. However, if a really recent upgrade was used on a really old previous version, data loss could occur.
The monthly column dealt with AUTOEXEC instructions and resulting memory use, and announced Beery Miller's "9640 News", a disk-based magazine.
The Q&A section dealt with some basic questions, such as what to do with the TI 32K card on a Geneve system (ditch it), what certain keys are for on the keyboard, and questions about the HFDC.
User notes included a way to load Funnelweb on the Geneve as if it were a cartridge (thus bypassing the need to load XB or EA into GPL first) and a way to change screen and character colors from the command line.
Jan 89
Genial Computerware sells HyperCopy by Mike Dodd. Most fast disk copiers wouldn't work on the Geneve, and the diskcopy command was slower than those programs. In addition, the program allowed formatting a disk with a skew, where the layout of a disk is shifted on each track, improving read times.
In the Q&A it was mentioned that device WDSx (Winchester DiSk) was being deprecated in favor of HDSx (Hard DiSk). The $75 upgrade involving adding an extra 32K of 0 wait state static RAM (to take it up to 64K) was mentioned Projects "on the board" included an enhancement to take the 12mhz Geneve up to 16 or 18 MHz. It was also mentioned that since 512K was the most RAM you could put on the PE Box expansion bus, the idea of a 1.5mb memory card was being dropped.
The features of MDOS 1.14 were outlined including:
- autoexec can be named anything you want and called from another file or the keyboard but the name must be preceeded by an ampersand when called (if the file were START you would type &START)
- **DESCRIBE TI FILE TYPES and use of DIR 'DV type switches
- Setting foreground and background to any of 16 colors with MODE Fx and MODE Bx
- Screen scrolling is faster
- WDS replaced with HDS
- You can pass filenames to GPL at the command line (GPL B:EA)
- CTRL ALT DEL exits GPL and returns to MDOS
Feb 89
Myarc announces that it will no longer announce release dates of forthcoming software.
Version 1.15 of MDOS has been finished but not released. MDM5 v 1.29 released (mostly dealing with backing up hard disks).
Jeff Kittka published the Pallette Master program used to mix colors. Only 16 colors are available in TI Extended BASIC, but this program allows those 16 to be defined as any of the Geneve's own 512 colors, rather than being stuck with the 4A's original palette (you could, for example, get a proper flesh tone or brown, instead of shades of yellow and orange).
The Q&A discussed creating directories on floppies, explained the GEME only allowed running 4 simultaneous programs because more than that would slow the system, various monitors for use with the Geneve.
Mar 89
Myarc posts beta copies of Advanced BASIC to bulletin boards.
The Q&A suggested users could change the font in MDOS with a sector editor, although it didn't elaborate. The C compiler was mentioned again (after several months) and it was revealed that this was under development by Pecan and would run under the much-delayed (and ultimately released as a limited “last ditch” version) Pascal Runtime environment. Myarc's plans to develop a relational database were formally cancelled, although Riley mentioned that a database called Swan would likely run under Pascal Runtime.
Apr 89
Two very different reader letters were printed - one complaining about the lack of response from Myarc, and the other a glowing review of their service.
A software listing for XDIR by John Johnson was printed, that let you run a directory with various options, including pausing, printing, and displaying only certain file types.
May 89
Numerous letters appeared defending Myarc and its support.
July 89
Advanced BASIC beta testing is nearly complete and "minor problems" remain.
A reader from England noted that the May letters supporting Myarc were all from the US, and noted that "overseas supporters are having a bad time of it."
A patch for Archiver 3 was printed, as was a series of batch files to allow easy operation of the Geneve.
  • Like 11
Link to comment
Share on other sites

Hmm...found this version, too. Bit less "timeline-ish."

 

The Myarc Geneve 9640 Family Computer

HISTORY

A story in the first issue of "Home Computer Compendium" (soon after known

as "microPendium", Texas Instruments having objected to the use of their

trademark "Home computer") in February of 1984 announced the "99/64 (aka

Phoenix)" from CorComp. CorComp had already established itself as a maker

of memory, disk controller, and RS232 cards that worked with the 99/4A's

Peripheral Expansion System.

Features of the announced machine included 64K of RAM exapandable to 1mb,

built-in RS232 and disk controller peripherals, an improved Extended BASIC,

and up to 132-column display. Compared to the 32K (max) RAM, expensive

peripherals, a somewhat clunky but powerful Extended BASIC, and 32/40-column

display, this machine was called a "Dream Machine" by the article.

How true.

It turns out that the details would be only partially realized. CorComp was

to release an expansion module that included these features but nothing

more. No new computer. No built-in "super BASIC". No improved display.

This device, the Micro Expansion System, was designed for those 99ers that

had not yet purchased TI's bulky Peripheral Expansion System.

However, it was evident that 99ers were hungry for an upgrade path not even

a year after TI abandoned the Home Computer market in Autumn 1983 that

wouldn't waste their previous investment in hardware and software. After

TI left the market, user groups in several cities, especially Boston and

Chicago, held "TI Faires" and it was at such events that new products for

the orphaned 99/4A were usually announced.

ENTER MYARC

Myarc (Microcomputer Architects), like CorComp, had its beginnings in the

production of memory expansion, RS232, and disk controller cards. They

expanded their line with an improved BASIC and RAMdisk cards.

At the March 1984 TICOFF exhibition in New Jersey, Lou Phillips, owner of

Myarc, announced a new 4A compatible computer. This machine was announced

in two versions – a self-contained model with keyboard (similar in design to

the 99/8 or early Amiga and Atari machines of the time) and a

computer-on-a-card that would fit in the TI Peripheral Expansion System.

The PES (or "P-Box" as it's known) connected to the 4A via a "Flex Interface

Card" and a huge cable that so resembled a fire hose that it was often

called just that.

Further details of the as-yet unnamed machine were given the next month, at

the New England 99 Faire in Boston on April 5, 1984. 256K or 512K RAM

(expandable to nearly 2mb), a TMS 9938 video chip (later produced by Yamaha

in many MSX and other computers) that could handle 80-column text and

graphics at 512x424x256 colors, RGB and composite video output, a mouse

port, PC keyboard interface, SN76496 sound chip (compatible with the sound

chip in the 4A), and TI's TMS 9995 CPU – a 12mhz successor to the 3mhz

TMS9900 used in the 4A.

WHERE'S THE COMPUTER?

By 1985, the Chicago TI Faire had become the premier event in the TI world.

The November 2nd show saw 1,700 visitors expecting to witness the debut of

the Myarc computer.

What they saw instead was an Amiga-style computer case with built-in

keyboard and a cartridge slot. Lou Phillips was announcing the machine

would start shipping in the first quarter of 1986 at a price of around $450.

That first quarter came and went.

The vaporware situation was becoming so frustrating for 99ers that in early

1986, microPendium changed its masthead from "Covering The TI99/4A Home

Computer And Compatibles" to "Covering The TI99/4A EXCLUSIVELY".

THE GENEVE COMETH

Myarc finally announced that the computer would ship in card-only form and

would be named the Myarc Geneve 9640 Family Computer (more commonly as

either "the 9640" or "the Geneve" by users). Rumor has it that they wanted

to use "99" in the name, but TI wouldn't allow it. The "9640" came about as

the "9" refers to "99/4A" and "640" refers to the RAM standard in the

machine. "Family Computer" was a nice was around TI's trademark of "Home

Computer." And, feeling that they needed a "friendly" name for the machine,

Lou Phillips and Jack Riley of Myarc saw "Geneve" on a painting and went

with that.

Fancy, glossy, color flyers were mailed out to thousands of 4A owners as

Myarc apparently bought TI's own mailing list (made of names of those 4A

owners who returned registration cards to TI after their purchase).

In February 1987, microPendium again changed its masthead to "Covering the

TI99/4A, the Myarc 9640 and compatibles". The 9640 was to be the only

"compatible" ever produced.

The first formal review of the machine was in microPENDIUM's April 1987

edition. This was reviewing a beta-test machine. The article starts, "The

Geneve 9640 is here! Finally. And it works."

A production model was received the next month, although the May 1987 issue

states that the editors were "still awaiting release of M-DOS."

SOFTWARE

Several pieces of software were announced with the Geneve, including:

1. Cartridge Saver – Most of the software produced during by TI during the

production run of the 4A was in cartridge format. Since the Geneve lacked a

cartridge port, a program that ran on the 4A allowed a 4A user to dump

cartridges to disk in a format that would work with the Geneve. This format

turns out to be the same as that used with the GRAMkracker – a cartridge

port device for the 4A that also allowed cartridge contents to be saved to

disk and, optionally, manipulated (such as changed default filenames when

saving data to disk with some programs, or changing default printers from

the serial port TI preferred to the parallel port which everyone else

preferred).

2. Advanced BASIC – this was a rewrite of Myarc's own Extended BASIC II and

was for a time known as Extended BASIC III. It's largely compatible with 4A

BASIC and Extended BASIC, although it allows access to the Geneve's features

such as advanced video modes.

3. 4.21 Pascal runtime – The 4A has a USCD Pascal card that allowed it to

run USCD Pascal programs. In essence, this card gave the 4A and additional

Pascal-based operating system. While powerful, the user base of the 4A

included mostly computer novices and families who had little use for such a

device.

4. TI Writer upgraded to 80 columns – The de facto standard in 4A word

processing was TI Writer largely because it was released by TI into the

public domain when TI left the Home Computer market. TI Writer was largely

based on the text editing system used on TI's 990 minicomputers. Instead of

WYSIWYG display, "dot" commands were entered within the text to produce

things like bold text or centered lines, similar to those used by the

WordStar program found on other micros. The text editor was used to edit

text, and a separate formatter program was used to interpret the dot

commands and print formatted text. The principal limitation was the 4A's

own 40-column display. For years, and 80-column display was the Holy Grail

of 99ers, and this modification of TI Writer allowed it to work in 80-column

text mode. Additionally, whereas the 4A's computer made entering commands

tedious (as well as text…FCTN P was the keystroke used to type simple double

quotes!), the Geneve's IBM-style keyboard was far easier to use for word

processing. Dedicated function keys made this particular task even simpler.

5. Microsoft Multiplan Upgrade – Since Multiplan was still a copyrighted

program by Microsoft, it could not be included in its entirety. However,

owners of the 4A's version of Multiplan (which included a cartridge and a

disk) could use Cartridge Saver on the cartridge and this patch on the disk

to get a Geneve-compatible version that ran faster and in 80-column mode. A

236 cell sheet took 2 minutes and 18 seconds to recalculate on a 4A, but

only 23.73 seconds on a Geneve.

6. Myarc Disk Operating System – The 4A had no "real" DOS. Although BASIC

programs and data could be loaded from and saved to disk, and programs could

be written to obtain a disk catalog, all other DOS functions (formatting

disks, copying files, etc.) had to be achieved with a Disk Manager program.

MDOS changed that, with a command-line and syntax very much like Microsoft's

MS-DOS.

AND THEREIN WAS THE PROBLEM

Myarc was under pressure from two fronts concerning the Geneve. First, they

had announced the machine 2 years previously and announced in December of

1986 that production had started. They really wanted to get the machine

into production as did their retailers, who had been taking advanced

payments – a practice that stopped when production delays plagued the

machine.

Second, they had spent a lot of money in developing this machine and had

very few resources (perhaps 5 employees and a couple of contractors)

designing it and the company had to start recouping its investment.

When released, it was painfully obvious that the Geneve was rushed.

Although the MDOS manual indicated users would see an "A:" prompt, the very

first pre-release featured a "DSK1." prompt. "DSK1" was how the 4A referred

to disk drive 1. A user would launch a program by typing its name from this

prompt. Most MDOS commands mentioned in the manual were not yet

implemented.

In fact, what Myarc was calling MDOS was actually its GPL Interpreter. GPL

– or Graphics Programming Language – was a TI-invented language somewhat

akin to assembly. Most cartridges were programmed in GPL, and it was this

interpreter that allowed a Geneve owner to run the cartridges they had saved

with Cartridge Saver. So while the Geneve could run most 4A software it

could run little else.

The first official Geneve column in microPENDIUM was a September 1987

article by Mike Dodd that largely talked about the various pieces of

software meant to come with the machine, and what the status was of each.

Finally, MDOS v.97 was released in October of 1997 and incorporated most of

the commands from the manual that shipped with the computer. v.99 added

batch file processing. Updated software was distributed on numerous

information services (such as CompuServe) and from time-to-time disks would

be mailed from Myarc to registered Geneve owners.

It was expected that a boot ROM would be released containing the "final"

version of MDOS so that users would no longer need to boot from disk.

However, the fact that MDOS was incomplete shelved this idea.

Ultimately, version 1.0 of MDOS shipped around December of 1987. GPL

Interpreter was at version .98 and ran from within MDOS. With the exception

of 6 minor differences, MDOS was now the same program described in its

manuals.

APPLICATION SOFTWARE AND "MY-ART"

Shortly after the Geneve's release, Myarc let fly a series of announcements

on software. They blamed the production delays mostly to the complexities

of the Geneve hardware and the need to write MDOS from scratch. They

claimed to have a system in place whereby certain pieces of PC software

could easily be ported. Early announcements included My-Number (a Lotus

1-2-3 work-alike), My-BASIC (a BASIC compiler), and My-Data (a dBase III

clone), none of which were ever released.

Futher vaporware included a C Compiler (considered critical since so much

software was written in C, and this would allow a great amount of software

to be ported to the Geneve) and My-Word Pro, an advanced graphical version

of My-Word that would support the Myarc Mouse.

However Myarc also announced, and actually released, its first and only

standalone application package. This was My-Art, a drawing program

retailing for around $149 and including the Myarc Mouse.

HFDC

Hot on the heels of the Geneve was Myarc's announcement of the Hard and

Floppy Disk Controller. This card allowed up to 3 134mb MFM hard drives to

be used with either a 4A or the Geneve. A streamer tape backup port was

also on the card, but never really worked.

It was therefore common for users who could afford a Geneve to also have an

HFDC card, and enjoy an added speed boost in booting MDOS and loading

software from hard disk instead of floppy disk.

Although older disk controller cards could make use of 720K 3.5" floppy

drives, the HFDC allowed the use of 1.4mb (high-density) 3.5" floppies as

well.

A special version of MDOS was created to work with the HFDC. For several

years, you had to use an "H" version of MDOS if you wanted to use a hard

drive or an "F" version if you wanted to use floppy drives. Neither version

would utilize both types of drives, and a few bugs remained in MDOS.

SPEECH

The 4A supported its Speech Synthesizer via the expansion port found on the

right side of the 4A console. This is the same port used to connect the

Peripheral Expansion System (and its "fire hose" cable). The synthesizer

had a pass-through connector, so most consoles had speech attached to the

console, and the fire hose plugged into the pass-through port of the

synthesizer.

Of course, the Geneve had no such expansion port. A company called Rave 99,

best-known for a keyboard interface that allowed PC-type keyboards to be

attached to the 4A, developed a "speech adapter card." This card allowed a

user to remove the Speech Synthesizer from its housing, plug it in the card,

then plug the card into the TI expansion system.

This remains the only way to get speech on the Geneve.

ADDITIONAL APPLICATIONS

Few pieces of software were released that ran natively from MDOS. The

Printer's Apprentice by McGann Software, was an advanced desktop publishing

package that ran on the 4A but was so complex that few 99ers could figure it

out. An MDOS version was released that utilized the Geneve's graphics and

memory.

TRIAD was a package that bundled a text editor, terminal emulator, and disk

manager.

A collection of games that were originally released on the TOMY TUTOR

computer were also released. These ports were made simpler by the fact that

the TUTOR shared much of the hardware found in the 4A and Geneve, including

the Geneve's 9995 CPU.

Much of the better TI software was written in Assembly language. Although

some of this software included Extended BASIC loaders, a few had to be

loaded using TI's Editor/Assembler cartridge. While Extended BASIC would

autoload a program on the first disk drive if named "LOAD", E/A had no such

niceties, and programs could be notoriously difficult to launch. An E/A

Program file loader was released to allow a user to launch these programs

from the MDOS command line. The only other way to do it was to boot MDOS,

load the GPL Interpreter, load in the Editor/Assembler cartridge, then load

the desired program in the usual way – adding burden to an already

cumbersome process.

ENTER BEERY MILLER

Beery Miller launched 9640 News, an on-disk magazine with the debut August

1988 issue. In 1990, he released three significant pieces of software –

Baricade (a game running out of Advanced BASIC), Tetris (which ran directly

from MDOS and is one of the few games to do so), and Windows 9640.

By the time MDOS was finalized enough to be really usable, the GUI market

was heating up in the rest of the computing world. The Amiga, Mac, Atari

ST, and DOS programs like OS/2, GEOS, and of course Windows, were proving to

the world that graphics and a mouse would make computing simpler for users.

Beery released Windows 9640 as a response. Although lacking in the

graphical prowess of the aforementioned systems (developed by teams of

programmers), it allowed task switching of up to 8 programs.

A similar program was under development by Myarc at one time. Named GEME,

it was completed and released by Beery with Myarc's permission in November

of 1991.

9640 News shut down in 1999

ENOUGH IS ENOUGH – THE BUYOUT

By the middle of 1992, development of MDOS had stalled. Paul Charlton was

the developer of MDOS for Myarc and after a dispute with the company,

refused to do any more updates to MDOS or release the source code so they

could update it themselves (this is a good reason to have clear ownership

and release provisions in software contracts).

Beery founded and facilitated a program to buy out the source code to MDOS.

The deal was that Beery would gather the funds ($XXX) for Paul and that Paul

would release the source code to Beery. When enough supporters chipped in,

Beery flew to New York, met with Paul, gave him the money, and was

personally handed the source code to MDOS. At long last, MDOS was in the

hands of the community with source code freely available. Versions 5 and 6

of MDOS were direct results of Beery's efforts.

HARDWARE SUMMARY

• TMS9995 CPU as 12mhz

• TMS9938 (Yamaha V9938) 128K display chip (512x424x256 colors, 80-column

text)

• TMS9901 Interrupt Controller

• SN76496 sound processor

• MM58274 real-time clock

• 512K 1-wait-state CPU RAM (expandable to 2mb)

• 32K 0-wait-state CPU RAM (expandable to 64K)

• The on-board 512K RAM can be disabled so 0-wait-state RAM can replace it,

but this renders the 99/4A emulation mode useless

PORTS

• Video – 8-pin DIN, 5 pins of which are the same as the 99/4A console to

allow use of a 4A owners existing composite monitor (albeit at lower

resolutions). The remaining 3 pins carry RGB blue, RGB green, and RGB sync

signals. The RGB red signal is activated via a jumper on the Geneve board,

thus turning the 5-pin composite port into an 8-pin RGB port. The video

port carries a composite sound signal, and most Geneve monitor cables

include a 1/8" plug for standard computer speakers or headphones.

• Mouse – Uses 9-pin mice.

• Joystick – This is a TI-compatible 9-pin joystick port. Adapters can be

plugged in to allow use of more-common Atari-compatible joysticks

• Keyboard – Allows PC XT keyboards to be used.

SCREENSHOTS

(Captured by Andy Frueh unless otherwise noted on the image).

Normally, the initial screen for the Geneve is a high-res image of a swimming swan (“the ugly duckling” being the TI-99/4A…get it?) However, a popular boot ROM let you hold down the space bar for a menu of devices from which to boot (floppy, hard drive, battery-backed RAM disk card, etc). The screen use 40-column graphics mode.

This is the initial (80-colum) screen that appears if the Genve’s built-in clock/calendar is unable to discern the current date and/or time. Later versions of MDOS are Y2K-compliant!

If MDOS finds an AUTOEXEC file in the root of the boot volume, it will execute the commands within, just as MS-DOS does with AUTOEXEC.BAT files.

Showing the MS-DOS similarities of MDOS, this is a sample DIR(ectory) command. Note that the hard drive is drive E:. This is because all disk controllers support at least three floppies, and the Myarc Hard and Floppy (HFDC) card supported 4. Thus, A through D were held in “reserve” making E: the first letter available for hard drives.

Alternatively, you could refer to the hard drives as HDS1 and HDS2, and floppy drives as DSK1 through DSK4. MDOS accepts either a period or a backslash in path names.

The use of either drive letters or device names and slashes or periods was a wise move to bridge the gap between TI users (device names and periods) and those familiar with MS-DOS.

The splash screen (again in lower-res mode for some reason) for the Windowing program GEME. One of the many pieces of software that Myarc started but ultimately never finished, this was meant to be a quasi-GUI that would allow you to run multiple programs in windows on the screen – somewhat akin to GEM, Atari’s TOS, DOSshell, and similar programs.

Not all programs had such boring displays. Here’s a game written in Myarc Advanced BASIC (ABASIC) that uses a high-res photo and line art for its splash screen (the remainder of the game is in 80-column text mode).

The GPL Interpreter allowed you to run cartridge-based software for the TI-99/4A that you previously saved to disk format with the CSAVE utility. The format of the resulting disk files is the same as that used by the TI accessory GRAM Kracker. Files saved on a TI with the “GK” will usually run fine out of the GPL Interpreter.

GPL speed is set with the F4 key. A setting of 1 is TI (slow) speed, and is often the required speed for any TI games that use speech. The faster settings are ideal for programs like spreadsheets and word processors.

The F6 option is unique…a lot of TI software expects to be in the first disk drive, DSK1. With DSK1 emulation “on”, you could create a subdirectory on your first hard drive called DSK1 and place files in it. The emulated GPL/TI mode would then see that directory as if it were the actual DSK1 floppy disk drive. CTRL SHIFT SHIFT would usually (but not always) take you out of the GPL environment and back to this loader screen.

The user interface for My-Word, the included word processing software. It is largely based on TI-Writer (which itself was largely based on text processing systems found on the TI 990 minicomputers) which was placed in the public domain by TI upon leaving the home computer market.

The program uses dot commands for formatting very similar to what Wordstar did on the PC. For example, placing .C centers text while .B turns on bold face. Text files using such formatting must be run through a separate “Formatter” program (as shown in the menu bar) to “interpret” the dot commands into actual print. There is no WYSIWYG mode, although the later aborted Press program from Asgard Software was working on a graphical word processor for the TI and Geneve.

A Help file is included.

In these days, programs just weren’t programs without having a windowing menu/interface. Shown are Config Your Autoexec by Tim Tesch, which aids in the creation of AUTOEXEC files, and Directory Manager by Clint Pulley, which is a nice directory/file manager program useful on the command-line oriented Geneve.

Here are a few sample drawings from My-Art, the drawing program sold (or “included free” if you prefer) with the Myarc Mouse. They show the graphics potential of the Geneve – certainly akin to the Ataris and Amigas being produced at the time (if not slightly better) and far exceeding the capabilities of PCs and Macs at the time.

Though these games are all low-resolution (ported from the TOMY Tutor, which was basically a TI-99/4A console!), they show some of the potential of the machine in lowe-res mode. And yeah, two of these are apparently clones (Scramble and Pooyan).

The Myarc Geneve “computer-on-a-card”. This card would be inserted in the TI Peripheral Expansion System in slot one – the slot normally taken up by the Flex Interface Card (the “firehouse”) that connected to the 4A console. The 4A console was unused in the Geneve system. Any memory expansion cards that gave the TI its additional 32K of RAM had to be removed or altered, as they would interfere with the Geneve. RS232 and Disk Controller cards were generally compatible with the system.

It could be said that the never-released TI-99/8 was the predecessor to the Geneve. It shared the same processor and speed characteristics. However, it had a cartridge port (not 4A compatible) and featured the new TI Hex Bus interface – a simpler and cheaper way to daisy-chain peripherals vs. the Peripheral Expansion System. In many ways – such as the DOS, the graphics – the Geneve was a superior machine to this never-released successor to the 99/4A.

  • Like 3
Link to comment
Share on other sites

HISTORY

 

[...] and DOS programs like OS/2 [...] were proving to the world that graphics and a mouse would make computing simpler for users.

 

This is seriously hurting a dedicated OS/2 fan like me. (Yes, I know, it's not from you. :) )

Link to comment
Share on other sites

 

This is seriously hurting a dedicated OS/2 fan like me. (Yes, I know, it's not from you. :) )

 

Ha, doesn't mean I can't edit the error out. I run three different versions of OS/2 at home. Long time fan here, too.

 

I also noticed that somewhere I got the info that the 99/8 cartridge port wasn't 4A compatible. The 99/8 manuals and technical documents from TI say otherwise.

Link to comment
Share on other sites

The 99/8 cartridge port does have some differences, as I had to learn when I recently reimplemented the 99/8 emulation for deeper emulation detail.

  • The 99/8 grom port has no CRUIN line (pin 6 is NC)
  • The 99/8 grom port has an A2 line as pin 4, which is CRUCLK on 99/4A

This additional address line means that the 99/8 port can handle cartridges with 16 KiB ROM space. It will not run any cartridge with CRU devices, but all others should be supported.

  • Like 1
Link to comment
Share on other sites

The 99/8 cartridge port does have some differences, as I had to learn when I recently reimplemented the 99/8 emulation for deeper emulation detail.

  • The 99/8 grom port has no CRUIN line (pin 6 is NC)
  • The 99/8 grom port has an A2 line as pin 4, which is CRUCLK on 99/4A

This additional address line means that the 99/8 port can handle cartridges with 16 KiB ROM space. It will not run any cartridge with CRU devices, but all others should be supported.

 

Do you know which cartridges would not run? (maybe it is interesting for Klaus and Ralf to list it?)

 

(Similar as for the /4 some cartridges cannot run, because of the video-chip and I belief Ksarul mentioned

that some cartridges cannot be inserted properly, because of the port design or dimensions are slightly different)

Edited by globeron
Link to comment
Share on other sites

Pin 4 of the cartridge port is not connected at all on QI motherboards either, and possibly on some of the last ones before the QI boards came out. . .something Marc Hull and I discovered by actually trying to use that signal, only to find out that it wasn't always there.

  • Like 1
Link to comment
Share on other sites

  • 4 years later...

On the Geneve clock: MM58274

I wonder if it has this trimmer

Crystal Oscillator
This consists of a CMOS inverter/amplifier with an on-chip
bias resistor. Externally a 20 pF capacitor, a 6 pF –36 pF
trimmer capacitor and a crystal are suggested to complete
the 32.768 kHz timekeeping oscillator circuit.
The 6 pF –36 pF trimmer fine tunes the crystal load imped-
ance, optimizing the oscillator stability. When properly ad-
justed (i.e., to the crystal frequency of 32.768 kHz), the cir-
cuit will display a frequency variation with voltage of less
than 3 ppm/V. When an external oscillator is used, connect
to oscillator input and float (no connection) the oscillator
output.
When the chip is enabled into test mode, the oscillator is
gated onto the interrupt output pin giving a buffered oscilla-
tor output that can be used to set the crystal frequency
when the device is installed in a system. For further informa-
tion see the section on Test Mode.

  • Like 1
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
 Share

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...