TMS9900 assembly language

[Bones]

Has anybody done any assembly language programming on a TMS9900 microprocessor? I have been crunching through some books (up to my third one now), and the pieces are just not falling into place.

Am I THAT stupid or is it really THIS hard?

[ChadTower]

Best guess - you're working on a TI99?  I've done a little - and I do mean a little - TMS9900, but it was years ago.  I think I was like 10.  I know you're a fellow TI99 guy... IMO, it's harder now than it was then because back then it was cutting edge - it was proper technique.  Nowadays you get guys with masters degrees in cmpsci that have never seen assembly.

IMO, the biggest problem with assembly on the TI99 was that the CPU was 16 bit but the damn RAM was only 8 bit... it crippled the instruction set.  That's why nearly all programming for the TI99 was done in basic... though look into GPL, which IIRC was an interpreter that was easier than TMS9900 and could be run on top of the CPU on the TI994/A.

Of course, if my guess of your motivation is wrong, slap me in the head and tell me to move on.

Also, I'd be surprised if there aren't any "post modern" compilers for that CPU... maybe a C compiler for it now, written well after the fact by fans.

[Bones]

Best guess - you're working on a TI99?

Yep, you are spot on.

I just found the original Editor Assembler documentation online and ordered it. Hopefully this might offer a little more enlightenment....

[ChadTower]

You just want to say naughty words through the speech synthesizer, don't you?

[SavannahLion]

Not to steal the topic here.

Nowadays you get guys with masters degrees in cmpsci that have never seen assembly.

It's worse than that. Some universities are churning out largely incompetent people.

I had to deal with a professor who was adamant about doing "proper techniques." In a nutshell, this meant writing code, regardless of system impact, efficiency or maintainability. So his idea of "correct" code was something that took up thirty or so lines wwhen the same practical code could be condensed down to ten or fifteen.

I understand the need to teach computing theory, but teaching the intricacies of a Bubble Sort for three weeks and spending only one week, total, on other sorts is just plain sadistic.

Another professor have the gall to tell me, "Don't ever worry about wasting RAM, there's always plenty." At the time I had to use a borrowed PC with only 64MB.

Another always required all variables, "at the top," forcing all to be global. (Turns out she checked the, "validity," of the code by running the code once, then comparing the variables to her "master" program.  )

There should be an exit course required for all CompSci majors, Practical Application 101.

And now, we return to your regularly scheduled topic.