• Please review our updated Terms and Rules here

Dvorak alternative history?

geoffm3

Veteran Member
Joined
Oct 9, 2009
Messages
1,272
Location
Huntsville, AL
There's some interesting... ahem, anecdotes in this article.

http://www.dvorak.org/blog/whatever-happened-to-the-texas-instruments-home-computer/

While there's certainly some good points to be made for the lack of certain keys on the keyboard, I do recall writing a few papers with TI Writer with little trouble, a lot of the information in here is counter to what I've heard the major issues were with the 99/4A:

1. the safety recall which wiped out profits from the machine ("Safety Checked")
2. the price war with Commodore
3. the high cost of manufacturing compared to rivals, which is odd considering it's even more vertically integrated than Commodore's manufacturing process was. I suppose a lot of that had to do with the fact that the 9900 was less of a commodity item than say a Z80 or 6502 processor and the insistence of upper management to use only TI parts.
 
The keyboard just wasn't that much of an issue. If it was, you can bet there would have been more aftermarket keyboards for the 99/4 than the Commodore PET. Were there even any at all?

The biggest issue was speed. It was slower than molasses. The second biggest issue was cost. By the time you got a cassette drive hooked up, you could have bought an Apple ][+ and Disk ][.

The 9900 was a poor design, and the 99/4 in an attempt to water down the 9900 even further, was very crippled.

TI made two big mistakes in my opinion: First they built a theoretical machine rather than a practical one. Second, they did something that was real popular everywhere except at Commodore (until Tramiel left!); they thought they could sell a bigger, better one by purposely selling a lesser unit first.
 
The TI 99/4a was a game console more than computer, and the Atari 400 games were much better than anything TI offered. There were features in TI-BASIC documented in the Manual that were never actually implemented in the computer. I forget if it was a rewind or backspace function for tape operations that passed through the interpreter but did nothing. TI was not going to sell a 16-bit computer for that price that was going to take sales away from their business machines. The assembly language for the 9900 was nice; the instruction set was powerful; but you just could not do anything with it from the computer.

The TI CC-40 and HEXbus peripherals- CC-40, read about all the promised HEXbus products in the pipeline, but the computer- use it for one application. The age of "Vaporware". It was an 80C85 processor. TI seemed good at components, good at calculators, but computers- just did not provide the support required to make a computer system work.
 
Not good at calculators either. At the same time as the TI 99/4A, TI also was working on the TI-88 which was slow and buggy. It got canceled and the factory was turned over to increase TI 99 production. So not only did TI lose even more money on the 99, TI also managed to disappear from the high end programmable calculator market for a few years. Not a good time for those of us who didn't want to buy from HP.
 
The keyboard just wasn't that much of an issue. If it was, you can bet there would have been more aftermarket keyboards for the 99/4 than the Commodore PET. Were there even any at all?

The biggest issue was speed. It was slower than molasses. The second biggest issue was cost. By the time you got a cassette drive hooked up, you could have bought an Apple ][+ and Disk ][.

The 9900 was a poor design, and the 99/4 in an attempt to water down the 9900 even further, was very crippled.

TI made two big mistakes in my opinion: First they built a theoretical machine rather than a practical one. Second, they did something that was real popular everywhere except at Commodore (until Tramiel left!); they thought they could sell a bigger, better one by purposely selling a lesser unit first.

Aye, it was definitely slow, but I dunno about cost of a cassette system vs. an Apple II unless the 99/4 was different than the 4A. Our first "real" computer was a 99/4A (the "first" being a TS/1000). My dad bought a one before the big price drop for $310 at KMart and we used a portable cassette deck we had around the house at the time. This would have been about late 1982 or early 1983? Maybe there was a rebate too, not sure. I do remember my dad complaining that they offered the machine shortly after for substantially less. A complete system was astronomically expensive to be sure, but once you added 32K to the system it was considerably faster (although still not a speed demon).

Looking at some of the information on the Wiki page it sounds like they were planning for some parts that would have substantially reduced the parts count that never materialized in production.
 
Only a serious programmer would complain about the TI-99/4A's keyboard "missing keys". Having terrible flat membrane keyboards never stopped the Atari 400 and Sinclair ZX-81 from being huge sellers. For around the same price, TI gave you a "real" keyboard that you could actually touch-type on.

And Dvorak needs to check his eyesight, because as far as I am aware, no popular keyboard layout has ever had a dedicated question mark key! Typing a "?" is still a two-finger process regardless if it's Shift-/ or Fctn-I. Maybe the former was more familiar to existing computer users, but someone new to computers could just as easily get used to the latter.

All of the characters that the TI required you to hold down Fctn to press were normally Shifted on most other home computers of the time. And the TI gave you typable characters that many other early '80s computers lacked, such as braces, backslash, pipe, tilde, and grave accent.
 
The PET 2001 has a dedicated ? key. In fact, that keyboard is a BASIC programmer's dream. You almost never need to press shift at all.

I don't have a problem with the TI-99/4A keyboard. I don't remember the 99/4 keyboard well enough to comment on it. I could certainly get used to the 99/4A one if Ihad to. There were much worse keyboards. Actually, it has a halfway decent feel.

As a serious programmer, I have some serious gripes about the 99/4A. But the keyboard isn't one of them.
 
I had the 99/4a- Keyboard was fine, no complaints with it. You could even buy them at Radio Shack after the computer was long gone for hobby work.

The 99/4 had a "Chiclet" keyboard? I never had one, but played with one at a store- remember not liking it.

The SR-56 got me through calculus class. The teacher would take a numerical solution if you could write the code during the test period and turn in the programming sheet for the answer.
 
And Dvorak needs to check his eyesight, because as far as I am aware, no popular keyboard layout has ever had a dedicated question mark key! Typing a "?" is still a two-finger process regardless if it's Shift-/ or Fctn-I. Maybe the former was more familiar to existing computer users, but someone new to computers could just as easily get used to the latter.

Someone new to computers would be expecting a typerwriter keyboard layout, not function-I to produce a question mark (or function-B on the TI 99/4). A common use for a new home computer purchase was word processing, and I'd suspect "?" was a fairly common character. So I think the criticism is valid. Even the Apple II got this correct, years earlier.
 
I think there's a little bit of merit to the criticism, but I can't recall much said about it at the time.

FWIW, I thought it was a rather unfortunate thing that the Atari 8-bit machines would place the return key above the home row, but I see that my ADM-3a does as well.
 
The TI-99/4 did not sell for the retailers lucky enough to get some. Magazines may have been favorable but the general public wasn't. An inconvenient keyboard is more acceptable in a very cheap home computer than the $1,000+ TI thought the original 99/4 should command.
 
Someone new to computers would be expecting a typerwriter keyboard layout, not function-I to produce a question mark (or function-B on the TI 99/4). A common use for a new home computer purchase was word processing, and I'd suspect "?" was a fairly common character. So I think the criticism is valid. Even the Apple II got this correct, years earlier.

Statistically speaking, only /most/ contemporary typewriters "got it right". Anyone used to a typewriter was in for a surprise switching typewriters, let alone switching to computers.

The Apple ][ more resembles a teletype than a typewriter.

As a typist, Apple scored lower with me than TI for lack of lower case.
 
FWIW, I thought it was a rather unfortunate thing that the Atari 8-bit machines would place the return key above the home row, but I see that my ADM-3a does as well.

So did the Apple II and II Plus.

apple2keyboard.jpg


There really was no such thing as very good home computer keyboard back in the early '80s. Even the IBM PC's keyboard was highly criticized for its key placement, and the C64 wasted entire keys on uncommonly used symbols like £ and ↑ while it awkwardly doubled up the cursor keys as left/right and up/down. And try asking Brits how they liked the ZX Spectrum's keyboard... :p
 
Citation needed. I believe there were some issues with key size, but not placement.

People didn't like it having the backslash key between left Shift and A and the tilde key between the quote and Enter keys, causing accidental keystrokes if you didn't reach far enough for Shift or Enter. Although this layout did eventually become the norm in Europe (the ISO layout), IBM changed the placement of those keys on the AT's keyboard to resolve those complaints (further improved with the Model M and standardized as the ANSI layout).

Even today, a lot of people aren't happy with the ISO layout:

https://ultimatehackingkeyboard.com...r-iso-which-keyboard-layout-is-more-ergonomic
 
How early 80s do we get to compare? The B128 has an utterly awesome keyboard, except I hate it for one reason: the key layout is significantly different from the other Commodores.

Ten points to vwestlife for typing an up arrow without using a C64. That is slick! :D But I have to know, can you type the left arrow? It bugs me that I cannot, because my assembler makes extensive use of that symbol.
 
The 9900 was a poor design, and the 99/4 in an attempt to water down the 9900 even further, was very crippled.

The second half of this I fully concur with, but not the first. I always found the 9900 to be a very sane early 16-bit design with a decent number of registers and a solid instruction set. The ALU was a little weak, but it had full 16-bit data paths internally and out and was well-documented and easy to program for. TI really did the chip a disservice with the board implementation and the architecture didn't deserve that black eye.
 
I wish I had had the opportunity to see the TMS9900 in action in a proper 16-bit computer implementation.

The TI-99/4A hardware implementation was just insane. Imagine running a computer where all of your system memory was on your video card, and your video card was attached via a parallel port. Your only programming language for it is in an interpreter, which itself is implemented in an interpreter, and the ROMS that contain most of it are also connected to your computer via a parallel port. Internally, that's the TI-99/4a for you.

Only 128 bytes of true 16-bit RAM and a little bit of 16-bit ROM. And the above and expansion peripherals were dumbed down to 8-bit. So even if you had the 32k memory expansion, it was still only using probably at best half its speed potential. And then the expansion gets in to the whole mess of sidecar-type peripherals.

The small keyboard was nothing. Once you got used to it, it was fine. Trying to use it after being used to a full keyboard can be jarring (no dedicated backspace key!), and the keyboard overlays have usually long since walked off.
 
Back
Top