The one thing YOU think they should NOT have left out of the TI-99/4A.

[+Ksarul]

Core was a great persistent memory technology. It just took up too much space--and needed a lot of power too. You could do all kinds of crazy things with it too--just look at the wiring harness from the Apollo spacecraft. Wherever there was room to do so, there was a chunk of inline core memory. . .and it was flexible enough to allow you to put it wherever you needed it. That's one of my favorite exhibits at the Smithsonian Air and Space, along with the massive 8K core memory module they have tucked away in a corner of the same room.

[JamesD]

You could probably list everything from any TI wishlist or what they should of done thread.

The joystick ports and reset switch are definitely inconvenient, but lack of CPU RAM was crippling.
Even including something like 2K of scratchpad RAM would have been a huge improvement.

Someone needs to make up a TI-99 decision making chart. It should be filled with decisions like this:

a) Use common joystick port pinouts so owners can use existing 3rd party joysticks
b) Cripple it

and this:
a) Allow uses to expand CPU RAM internally and bank switch out the ROMs so users can really unleash the power of the machine
b) Cripple it

and this:

a) Include a finely tuned high speed BASIC that can use assembly language subroutines so programmers can get every last bit of performance out of the machine
b) Cripple it

and this:
a) Plan from day one on using an existing processor to reduce time to market and to allow the machine to use versions of existing software from the mini-computers on it which would give it access to a broad range of development tools

b) Cripple it

Now that we've completely crippled the machine, how do we make it sell?
a) Go back and fix these problems in the first revision
b) Make it talk

* Note, TI engineers must always choose "b"

[Stuart]

* Note, TI engineers must always choose "b"

 

I think that should be 'TI senior management/marketing must always choose "b"'. TI engineers left to their own devices developed the 9995-based Powertran Cortex, which includes several option "a"'s.

 

Stuart.

[Omega-TI]

 

I think that should be 'TI senior management/marketing must always choose "b"'. TI engineers left to their own devices developed the 9995-based Powertran Cortex, which includes several option "a"'s.

 

Stuart.

 

Awww... please cut the marketing guys some slack. I think you're probably right when it comes to senior management/accounting (bean counters), but the marketing/sales staff is a whole different matter. The latter would prefer to have the BEST quality product as it makes it easier to push units. Having to create a sales plan to overcome deficient, sub-standard or obsolete products is not an enjoyable experience and in many case requires ethically challenged individuals, which sometime can even taint a companies image.

 

I agree on the engineering staff, they are more "in-touch" with the cutting edge and know "what's cool". Being in constant daily contact with a developing project gives one insight into NEEDED features, which the out-of-touch suits in upper management may have no clue about.

 

To be fair, it was an emerging market and all the things we take for granted now as common knowledge were foreign concepts back then.

[+Lee Stewart]

I do not know how much truth there is to this, but I had heard that marketing for the TI Home Computer Division had been put in the hands of the engineering staff.

 

...lee

[Retrospect]

A Zilog 80A chip with 32K would have been a good alternative to the 9900 with 256 bytes.

 

There were two working prototypes that had these Z80's in them.

 

The computers were altered and access to anything hacky about the machine was ultimately denied.

[Tursi]

So it should have been an MSX?

 

Honestly that wouldn't have been so bad, I wish we'd had the MSX over in North America...

[+mizapf]

A Zilog 80A chip with 32K would have been a good alternative to the 9900 with 256 bytes.