Apparently TI was planning a TI-99/3, and possibly a TI-99/5

[matthew180]

I was digging through some older magazines on Archive.org and ran across a 99/4 review article, and there are suggestions that TI was talking about a smaller TI-99/3 system, and rumors of a more powerful TI-99/5 system.

 

It also says Microsoft wrote BASIC for the 99/4:

 

"TI's version of Basic has several peculiarities.  Written by Microsoft, it is compatible with TI's larger computer (the TI-990, which sells for about $10,000).  But it's quite different from Microsoft's other 5 famous Basics (PET Basic, Ohio Scientific Basic, CP/M Basic, Applesoft Basic, and TRS-89 Level II Basic); you'll have a hard time converting programs from those ''unfriendly 5'' to TI."

 

Anyway, I found the review interesting, especially from the perspective of engineers looking for computers, and how TI was only targeting the home market.

 

https://archive.org/details/bitsavers_DigitalDesN07197907_53495813/page/68/mode/2up

 

[Gary from OPA]

As far as I know the 99/3 became the 99/2 and there was going to be a 99/5 at some point but only a few prototypes were made and most likely it would end up becoming the 99/8 anyway.

 

https://99er.net/992.html

 

Quote

The TI-99/2 BASIC Computer was made to compete with the Timex Sinclair. The 99/2 only displayed black/white. It used a 16-bit high speed processor, 4.2K of memory, and a membrane keyboard usable by touch typists. The 99/2 could only run certain cartridges and tapes, the tapes gave a short introduction to the BASIC computer while the cartridges originally avalable for $19.95 each helped the user become familiar with BASIC. The cartridge names were, "Introduction to Programming" which plugged in the back of the console and immediatly turns the machine into an interactive teacher showing you how to communicate with it.

 

The other cartridge titled "Learn BASIC Programming" teaches you about programming in the BASIC launguage step by step.

 

The 99/2 BASIC Computer used a TMS9995 microprocessor operating at 10.7 Mhz. The 99/2 was made for families that were taking a step into the world of computing at a low cost. Then they would buy a more capable computer like the 99/4A (Color, sound, games, and tons of peripherals). The 99/2 was supposed to be a stepping stone to the 99/4A.

 

There were several ways to expand the TI-99/2. You could attach a shallow cradle on the bottom of the computer to add 16K - 32K-bytes of user RAM (for a total of up to 36.2K); an 8-pin connector on the rear panel of the machine allows connection of TI's Hex-bus family. There was also a Hex-bus compatible modem for the 99/2. Since the 99/2 and 99/4A can talk to the same peripherals over the Hex-bus, their data and program files can be interchanged through the Wafertape peripheral media. It might even be possible to "download" the 99/2 from the 99/4A through the Wafertape.

 

There was also an RS-232 interface, and a 4-color printer/plotter designed for the TI-99/2.

Basically the CC40 but no screen and using the tms9995 instead of the 7000 series. It should be emulated in MAME as well I think.

 

Most of the 99/5 information, is I think is in this AtariAge thread, I can't find any working current links to the pictures as the ones in this older thread seem to be dead now, but the person that owns one of the surviving units still is still around, he can post some updated online working links, etc.

 

 

 

Edited July 17 by Gary from OPA
more info and links, pics

[tschak909]

Can we STOP it with the "Microsoft wrote the BASIC for the 99/4?" That did not happen! The 99/4 BASIC is directly descended from the 990 BASIC, which was implemented from the ANSI spec.

 

Jesus, retro computing archaeology is rife with speculation transmogrified to fact. Sigh.

[Gary from OPA]

Some info on what was called 99/3b which used a z80 cpu instead was posted here:

 

 

[+OLD CS1]

7 hours ago, tschak909 said:

Can we STOP it with the "Microsoft wrote the BASIC for the 99/4?" That did not happen! The 99/4 BASIC is directly descended from the 990 BASIC, which was implemented from the ANSI spec.

 

Jesus, retro computing archaeology is rife with speculation transmogrified to fact. Sigh.

Easy, easy.  We have already fleshed this out in other threads thanks to the work of people like @FarmerPotato @Ksarul @acadiel and others.  No matter, it will turn up every now and then like this, due to pulling up misinformed articles from the time period.  At this stage, we can just let it flow over us like water over a duck's back.

 

(Now, tell me about how Windows 95 runs on top of DOS... might start a fight! )

[+mizapf]

8 hours ago, Gary from OPA said:

Basically the CC40 but no screen and using the tms9995 instead of the 7000 series. It should be emulated in MAME as well I think.

Most of the 99/5 information, is I think is in this AtariAge thread, I can't find any working current links to the pictures as the ones in this older thread seem to be dead now, but the person that owns one of the surviving units still is still around, he can post some updated online working links, etc.

I got the ROM dumps from Fabrice some years ago, it is on my list for MAME. I got distracted by SCSI, IDE, TIPI etc. 🙂

[+mizapf]

37 minutes ago, OLD CS1 said:

(Now, tell me about how Windows 95 runs on top of DOS... might start a fight! )

Ah, there's something more on-topic: Discuss whether the TI-99/4A is a "real" 16-bit computer. (Ah, no, forget it, my adrenaline starts to rise just by writing that line.)

 

[+FarmerPotato]

Browsing the 99/8's BASIC reference guide the other day, I noticed instructions for connecting two consoles by HexBus. Or maybe it was in the section about sending raw HexBus commands.

 

This would be a way to "download" from the 99/2.
The 99/8 manual hints at using the attached console to run a compute task.

 

Heck, the 99/2 with its 9995 could be a contributor if you turned off video! (VCHAR end of line chars in column 1.) Otherwise the 99/2 video steals too many RAM cycles. 

Imagine that some kid started with a 99/2 in 1983, then got a 99/8 and wondered what to do with that toy console. Reads an article in 99er about connecting the two...

 

 

 

 

[+FarmerPotato]

13 hours ago, Gary from OPA said:

Most of the 99/5 information, is I think is in this AtariAge thread, I can't find any working current links to the pictures as the ones in this older thread seem to be dead now, but the person that owns one of the surviving units still is still around, he can post some updated online working links, etc.

I had hands-on time with a console dubbed "99/5" by engineering manager Ron Wilcox. It's not quite the same as the 99/5 described by @fabrice montupet  I ran Parsec on it. It had Extended Basic II and proper RAM built in. 
 

Wilcox' 1982-83 papers show multiple prototypes of a 99/4B starting point, evolving into the 4A QI , 99/5 and 99/8.  A marketing document shows a strategy to add 2K console RAM , and to repackage command modules by minimum requirements. (Some would download GROM segments into the 2K.)

 

The point is, 99 model numbers were constantly redefined over 1981-83.
 

Don Bynum is active in reviewing and redirecting these efforts--particularly to roll the cost savings of 99/4B (and not its 2K extra RAM!) into the 99/4A QI.  Rebates and price wars shaved 4A margins so thin that mid-range 99/4B was not happening and of course the 99/2 became redundant.  Leaving just the high -end 99/8. 

 

 

I read recently in the TI Records at SMU that in March 83, the 99/8 was being positioned as a "Business Home Computer".  

That memo is out of OST, Office of Strategy and Tactics (higher-ups like Bernie List.)  It positions 99/8 and TIPC as separate market segments , with the 99/8 being more "business" than the 4A but still lesser than TIPC.
 

Another memo forecasts the retirement of 990s, recognizes 9900 as a disaster, and questions what DSG (Digital Systems Group) is going to do in the meantime. (NuBus, S1500 and Explorer were in development in 1982.)
 

 

The Mar 1983 paper on 99/8 and TIPC introduces the employees newly onboarded to 99/8 and TIPC.  It says that for business/home office, the 99/8 needs acceptably fast word processing etc, maybe database and bookkeeping?  Like the 990/1, they believed the small business would use BASIC and soon PASCAL. 
 

One employee was onboarded just to secure third-party software for 99/8. Maybe developed already in UCSD Pascal. 
 

Alas, history didn't happen that way. That outsider doofus, who hired Bill Cosby and stuffed the 4A channel with unsold consoles at Christmas 1982, showed a teeny profit that was soon erased.  (He got promoted higher than engineer Don Bynum.)  I guess when the true numbers got out, all of that Mar 1983 vision was disparaged. 

Read Wally Rhines' oral history at Computer History Museum. (He's the author of "9900: TI's biggest blunder".)    He bemoans that TI never promoted a semiconductor engineer to CEO and guessed why.  TI did not tolerate  a single failure from a head of engineering.    Wally coined the phrase "mean-time between crap-outs" to explain why no C-suite executive had ever run a Semi fab--fabs inevitably would have a a crap-out, and the chief was fired.  But Wally turned Microprocessors into TI's cash cow, before he left to become CEO of Mentor Graphics.  Maybe there were additional barriers to an engineer becoming CEO of TI.
 

Perhaps unwittingly, retiring CEO Pat Haggerty set up the TI Records archive to enable writing about TI's management philosophy over the years. He died before writing that book.  (Outsiders were hired to write "75 years of Innovation: Stories from Texas Instruments". Easy to find used copies for $10.)

 

Anyhow, I just wanted  to say that 99 varieties were constantly renamed during 1982-83.  By Mar 83 there was only the 99/4A and 99/8 in the strategy. 

 

The SMU archivist asked when I'm going to write a book. I said that I'm not that good a writer. 
 

 

 

[+OLD CS1]

3 hours ago, FarmerPotato said:

He got promoted higher than engineer Don Bynum.

The Dilbert Principle.

[tschak909]

2 minutes ago, OLD CS1 said:

The Dilbert Principle.

aka the William Sydnes Paradox. He went from designing the PCjr, to being the head of Commodore's engineering department from 1991 to 1994 (his first product out the door was the Amiga 600).

 

-Thom

Edited July 17 by tschak909

[Retrospect]

5 hours ago, mizapf said:

Discuss whether the TI-99/4A is a "real" 16-bit computer.

If a computer has an 8-bit CPU, it's an 8-bit computer.  If it has a 16-Bit CPU, it's a 16-bit computer no matter what the rest of it is.    

[tschak909]

4 minutes ago, Retrospect said:

If a computer has an 8-bit CPU, it's an 8-bit computer.  If it has a 16-Bit CPU, it's a 16-bit computer no matter what the rest of it is.    

for me, it's always been a case of what you can do with the basic maths operations, if you can add, subtract, etc in 16 bit chonks before you have to carry, it's a 16-bit machine.

[matthew180]

15 hours ago, tschak909 said:

Can we STOP it with the "Microsoft wrote the BASIC for the 99/4?" That did not happen! The 99/4 BASIC is directly descended from the 990 BASIC, which was implemented from the ANSI spec.

 

Jesus, retro computing archaeology is rife with speculation transmogrified to fact. Sigh.

 

I can't tell if that was directed at me, or the article?  I was just quoting the text, and certainly not making any claim or statements.  I know there has been a lot of discussion and research done in the area of the origin of TI BASIC.

 

But I find it interesting to see something that old (1979), and at the time when the 99/4 had just come out, still getting this confused.  And there is some half truth in there, citing that TI BASIC was compatible with the TI-990 BASIC, and *not* the same as the other BASIC dialects of the "unfriendly 5" (a term I have never heard).

[+FarmerPotato]

1 hour ago, matthew180 said:

And there is some half truth in there, citing that TI BASIC was compatible with the TI-990 BASIC,

I think @tschak909 brought in TI-990 BASIC. Never heard of that connection  being cited from the press.  

Its credible source is:

 

The CB Wilson papers gave us  a "Standard for TI BASIC" which you can see was mostly followed across TI products.  In retrospect, it proves the CC-40's Enhanced BASIC syntax was deliberate. 

[Retrospect]

I read the article Matthew referenced and my thoughts are that the journo was being a little speculative or just plain getting some things out of whack / wrong.  

 

The journalist only had what he was given to go on ... I mean he mentioned the graphics commands and left out GCHAR.  He wasn't what we could call knowledgeable.

[jbdigriz]

On 7/17/2024 at 1:10 AM, tschak909 said:

Can we STOP it with the "Microsoft wrote the BASIC for the 99/4?" That did not happen! The 99/4 BASIC is directly descended from the 990 BASIC, which was implemented from the ANSI spec.

 

Jesus, retro computing archaeology is rife with speculation transmogrified to fact. Sigh.

Bad enough we have to deal with what I consider to be official misinfo from the horse's mouth, so to speak, such as the IEEE "biggest blunder" article. 

 

It gets better...now we have GenAI models training on internet urban legends, ideology, unsubstantiated assertions, error,  and ignorance. Behooves us all to verify, document, and attribute our archeaology. If you look back on your old postings, you may wince, as I have. Trying to do better without maybe being too pedantic.  

Edited July 18 by jbdigriz

[jbdigriz]

23 hours ago, FarmerPotato said:

I had hands-on time with a console dubbed "99/5" by engineering manager Ron Wilcox. It's not quite the same as the 99/5 described by @fabrice montupet  I ran Parsec on it. It had Extended Basic II and proper RAM built in. 
 

Wilcox' 1982-83 papers show multiple prototypes of a 99/4B starting point, evolving into the 4A QI , 99/5 and 99/8.  A marketing document shows a strategy to add 2K console RAM , and to repackage command modules by minimum requirements. (Some would download GROM segments into the 2K.)

 

The point is, 99 model numbers were constantly redefined over 1981-83.
 

Don Bynum is active in reviewing and redirecting these efforts--particularly to roll the cost savings of 99/4B (and not its 2K extra RAM!) into the 99/4A QI.  Rebates and price wars shaved 4A margins so thin that mid-range 99/4B was not happening and of course the 99/2 became redundant.  Leaving just the high -end 99/8. 

 

 

I read recently in the TI Records at SMU that in March 83, the 99/8 was being positioned as a "Business Home Computer".  

That memo is out of OST, Office of Strategy and Tactics (higher-ups like Bernie List.)  It positions 99/8 and TIPC as separate market segments , with the 99/8 being more "business" than the 4A but still lesser than TIPC.
 

Another memo forecasts the retirement of 990s, recognizes 9900 as a disaster, and questions what DSG (Digital Systems Group) is going to do in the meantime. (NuBus, S1500 and Explorer were in development in 1982.)
 

 

The Mar 1983 paper on 99/8 and TIPC introduces the employees newly onboarded to 99/8 and TIPC.  It says that for business/home office, the 99/8 needs acceptably fast word processing etc, maybe database and bookkeeping?  Like the 990/1, they believed the small business would use BASIC and soon PASCAL. 
 

One employee was onboarded just to secure third-party software for 99/8. Maybe developed already in UCSD Pascal. 
 

Alas, history didn't happen that way. That outsider doofus, who hired Bill Cosby and stuffed the 4A channel with unsold consoles at Christmas 1982, showed a teeny profit that was soon erased.  (He got promoted higher than engineer Don Bynum.)  I guess when the true numbers got out, all of that Mar 1983 vision was disparaged. 

Read Wally Rhines' oral history at Computer History Museum. (He's the author of "9900: TI's biggest blunder".)    He bemoans that TI never promoted a semiconductor engineer to CEO and guessed why.  TI did not tolerate  a single failure from a head of engineering.    Wally coined the phrase "mean-time between crap-outs" to explain why no C-suite executive had ever run a Semi fab--fabs inevitably would have a a crap-out, and the chief was fired.  But Wally turned Microprocessors into TI's cash cow, before he left to become CEO of Mentor Graphics.  Maybe there were additional barriers to an engineer becoming CEO of TI.
 

Perhaps unwittingly, retiring CEO Pat Haggerty set up the TI Records archive to enable writing about TI's management philosophy over the years. He died before writing that book.  (Outsiders were hired to write "75 years of Innovation: Stories from Texas Instruments". Easy to find used copies for $10.)

 

Anyhow, I just wanted  to say that 99 varieties were constantly renamed during 1982-83.  By Mar 83 there was only the 99/4A and 99/8 in the strategy. 

 

The SMU archivist asked when I'm going to write a book. I said that I'm not that good a writer. 
 

 

 

I still have a bone or two to pick with Rhines about that IEEE article, but I agree he did a good job with the microprocessors. Wrt to the "mean time" phrase, I do recall posts on Delphi or some other early online forum that TI had a habit of chewing up and spitting engineers out in that era. (Some day maybe I will be able to resurrect those capture logs.) How true that might be though compared to other computer and semiconductor companies of the time, I'm not sure, but I've run across many former TI engineers and other personnel who were quite happy with their time there. Some of the "smart guys"  you mentioned in another post have gone on to even bigger things elsewhere, as Rhines did. I hesitate to mention names without at least touching base first. Sorry to violate my precept about documenting, elsewhere in this thread. 

 

Didn't know Haggerty had planned to write a book. Interesting. And thanks for reminding me I need to get back to that cataloging to submit to manx. Need to touch base with that archivist, too.

[Geoff Oltmans]

You would think if the Microsoft BASIC rumor passed any sort of litmus test, the first thing would be that it doesn't LOOK like MS BASIC.

[+OLD CS1]

9 minutes ago, Geoff Oltmans said:

You would think if the Microsoft BASIC rumor passed any sort of litmus test, the first thing would be that it doesn't LOOK like MS BASIC.

Well, as legend has it, TI demanded that it not look like MS BASIC

[Geoff Oltmans]

On 7/17/2024 at 2:13 PM, tschak909 said:

aka the William Sydnes Paradox. He went from designing the PCjr, to being the head of Commodore's engineering department from 1991 to 1994 (his first product out the door was the Amiga 600).

 

-Thom

Ironically, the Amiga 600 back in the day was much maligned as a step backwards from the Amiga 500. Fast forward to today and it's a more desirable model.

[+FarmerPotato]

4 hours ago, OLD CS1 said:

Well, as legend has it, TI demanded that it not look like MS BASIC

just legend?

 

Anyway here is one place to find the Jun 1978 Specification for TI Standard BASIC: this thread about a BASIC bug

[Gary from OPA]

51 minutes ago, FarmerPotato said:

just legend?

 

Anyway here is one place to find the Jun 1978 Specification for TI Standard BASIC: this thread about a BASIC bug

i read thru those threads, and it mentions this: (but I can't find any attachment pdf scan of it), was it ever completed? if so please attach it or give a link to it!

 

TI BASIC Interpreter System Documentation"....WRITTEN BY Robert B. Greenberg of Microsoft (8/24/1978).

 

So MICROSOFT is there... CASE SETTLED!

Edited July 24 by Gary from OPA

[tschak909]

2 minutes ago, Gary from OPA said:

i read thru those threads, and it mentions this: (but I can't find any attachment pdf scan of it), was it ever completed? if so please attach it or give a link to it!

 

TI BASIC Interpreter System Documentation"....WRITTEN BY Robert B. Greenberg of Microsoft (8/24/1978).

 

So MICROSOFT is there... CASE SETTLED!

Welp, Ric Weiland at Microsoft was approached by Atari to solicit a BASIC, so I guess that means Microsoft wrote Atari's BASIC.

 

Oh wait, that's not what happened.

 

-Thom

[oracle_jedi]

On 7/17/2024 at 5:43 AM, OLD CS1 said:

Easy, easy.  We have already fleshed this out in other threads thanks to the work of people like @FarmerPotato @Ksarul @acadiel and others.  No matter, it will turn up every now and then like this, due to pulling up misinformed articles from the time period.  At this stage, we can just let it flow over us like water over a duck's back.

 

(Now, tell me about how Windows 95 runs on top of DOS... might start a fight! )

 

Alas it is now baked into the AI LLM models that will soon run the world.  We are simply wrong, and this is now established "fact"