Really enjoying your TI99 videos. New to TI and getting into it now as of late. Have had one for a while but finally biting the bullet more as of late to learn more.
Great work overall! Really adds needed clarity to understanding just how easy the process of compiling BASIC on the TI has become. You even explained the Sprite Early clock so well! Bravo! I’d love to see a part two which explains modern editing via TICodEd. The compiler is great but TICodEd turns the whole deal into a modern SDK. Development workflows on modern PCs make programming on the TI even better and mobile.
I remember writing basic programs when i was 6 with the ti, my first computer. Nothing particularly useful, but it was such a sense of awe to even get things moving around the screen... Slowly.
So awesome! I really love your explanations on all this stuff. I also love your website. I’ve looked up many manuals for games I have. It’s really helpful and useful! Just entered the TI 99 community at the start of this year. It’s great to see that I’m not the only one who still loves this computer.
Glad the manuals site has come in handy! And glad you're enjoying your foray into the modern world of the TI-99. The luxuries of the modern Internet sure have done well by our little orphan platform.
Wow, I would have loved that if it was around in the early 1980s, back then I wrote various games in BASIC, but I never could have truly produced high quality commercial games without learning assembly like you can with this. I didn't learn assembly until considerably later on the Commodore 64 in 1985 as a teenager, then on the Atari ST and Amiga.
I recall buying a TI Compiler back in the day. For some reason, I could never figure out how to get it to work, in spite of the instructions. Then again, some 3rd party documentations were notorious for being bad. I so wanted to see my programs fly.
There were a couple attempts at it back in the day that got some kind of mail order or user group circulation, for sure. But in keeping with your experience, my impression is that they were pretty useless. Nothing even remotely comparable to Harry Wilhelm's product here, which is both utterly comprehensive in its support for Extended BASIC code and exceedingly well documented.
@@fitfogey : You’re so right! Blitz was a little iffy sometimes. If you had any errors in your code it would find them and puke. It took blitz about 20 minutes to finish compiling my BBS program. Those were the days!!
@@curiousottman Same here. Started off with a 300 baud modem on the C64. Over time eventually was able to get a 2400 baud screamer lol. I wrote/edited BBS code as well, had the second dedicated phone line, etc. They really were the good old days.
Ummmm... If the 4a had launched with 1GB of RAM at the same price point, it would absolutely been very successful, far more successful than Amstrad. It would _also_ out compete $500k workstations and many mainframes. So, beating Amstrad home computers would be small potatoes. :) Perhaps you meant 64k? I doubt that would have helped much. If you are looking for "ways to make the ti a market success," the answer is easy: expand the 256 byte "Scratch-Pad RAM" to, say, 16k. That, plus 32k of VDP RAM would produce an absurdly fast machine for the time.
@@jmr5125 Possibly a stupid question, but wasn't the "scratchpad" in-silicon on the CPU? I've always understood... possibly incorrectly... that this was the case. If this was an external chip, on the other hand, well... 256 bytes is what you get addressing byte-by-byte over an 8-bit address (as opposed to, say, addressing by memory words or the like). So, if the "scratchpad" was only addressed by an 8-bit address path, you couldn't improve that 256 byte region without reworking the CPU, anyway, regardless of whether the memory itself was internal or external. So, does anyone know the inner details of how this scratchpad region was really accessed? It seems logical, to me, that this would be on-die and would have been intended solely as a temporary holding space for intermediate calculation steps. That we try to use it for so many other purposes really comes down to how they skimped so terribly on REAL directly-addressable CPU memory in the system design. Had there been that 16-bit (64K max) CPU memory directly connected to the CPU over a dedicated memory bus... that would have been ALMOST as fast as scratchpad, and would have given us the performance that the system was, theoretically, capable of. But "Scratchpad" isn't really SUPPOSED to be a replacement for "directly addressable CPU memory." Sadly, it was the only such memory we got in the shipped design. And that was a design CHOICE made by someone in a high-level position at TI... probably someone who didn't want to undercut the market for their "mini-frame systems" by shipping similarly-performing units to the home market. In other words, it reads almost like internal sabotage... (sigh)
The VDP is capable of 15 different visible colours (and transparent). But the Standard Mode palette is divided into 32 palette groups (each of which may contain any combination of the available colours). So in memory, the Standard Mode palette is 64 values (4-bits each), each of which signifies a foreground or background colour for one of the 32 palette groups. So when you write those 32 bytes, you're setting 64 discrete values.
The F18A adds a great deal, but almost none of it can be exploited directly via BASIC. T80XB allowing 80 column mode via BASIC (with an F18A) is one limited example. I have to think someone will integrate a couple other key F18A features into whatever set of BASIC tools at some point, though. Or just as a standalone CALL LINK.
Very impressive but not really XB anymore is it? Hmmm RXB 2021 has ROLL up,down,left,right and SCROLL too. All are written in Assembly and will run from only a Console and RXB cart no 32K needed.
This is so well presented..I have tears 😭. But what a great addition to programming.
Really enjoying your TI99 videos. New to TI and getting into it now as of late. Have had one for a while but finally biting the bullet more as of late to learn more.
Great work overall! Really adds needed clarity to understanding just how easy the process of compiling BASIC on the TI has become. You even explained the Sprite Early clock so well! Bravo! I’d love to see a part two which explains modern editing via TICodEd. The compiler is great but TICodEd turns the whole deal into a modern SDK. Development workflows on modern PCs make programming on the TI even better and mobile.
The episode I've been waiting on....and it didn't disappoint! Great video!
Love the way the videos are presented.
I remember writing basic programs when i was 6 with the ti, my first computer. Nothing particularly useful, but it was such a sense of awe to even get things moving around the screen... Slowly.
This is a great presentation of a great XB tool for th TI!
So awesome! I really love your explanations on all this stuff. I also love your website. I’ve looked up many manuals for games I have. It’s really helpful and useful! Just entered the TI 99 community at the start of this year. It’s great to see that I’m not the only one who still loves this computer.
Glad the manuals site has come in handy! And glad you're enjoying your foray into the modern world of the TI-99. The luxuries of the modern Internet sure have done well by our little orphan platform.
Wow, I would have loved that if it was around in the early 1980s, back then I wrote various games in BASIC, but I never could have truly produced high quality commercial games without learning assembly like you can with this. I didn't learn assembly until considerably later on the Commodore 64 in 1985 as a teenager, then on the Atari ST and Amiga.
I remember the Tenex catalogs. I got Extended BASIC, Mini Memory, and Demon Attack from them.
I've put up colour scans of nine of them on my site, if you'd like to take a trip down memory lane: pixelpedant.com/items/browse?collection=35
Good Stuff!
I recall buying a TI Compiler back in the day. For some reason, I could never figure out how to get it to work, in spite of the instructions. Then again, some 3rd party documentations were notorious for being bad. I so wanted to see my programs fly.
There were a couple attempts at it back in the day that got some kind of mail order or user group circulation, for sure. But in keeping with your experience, my impression is that they were pretty useless. Nothing even remotely comparable to Harry Wilhelm's product here, which is both utterly comprehensive in its support for Extended BASIC code and exceedingly well documented.
I remember having a “compiler” for c64 basic called Blitz. It made my home brew BBS program run so much faster!
I remember that too. And when someone would sneaky try and list your program it would just say BLITZ!
@@fitfogey : You’re so right! Blitz was a little iffy sometimes. If you had any errors in your code it would find them and puke. It took blitz about 20 minutes to finish compiling my BBS program. Those were the days!!
@@curiousottman Exactly. Love the username by the way. 👍✌️
@@fitfogey : Thanks. Not many people get the modem reference. I miss the old dialup days.
@@curiousottman Same here. Started off with a 300 baud modem on the C64. Over time eventually was able to get a 2400 baud screamer lol. I wrote/edited BBS code as well, had the second dedicated phone line, etc. They really were the good old days.
I wish this guy could speak in assembly speed. I have to play the vid at 2x speed to not fall asleep lol
Cool!
Can’t imagine programming without a backspace lol!
That’s why you use the compiler with TICodEd on a modern PC.
Have you reviewed the TI Logo and TI Logo II?
I have not, but it's certainly something I've considered. Definitely one of the things on my list.
It was the first 16bit machine when only 8bit machines were around.
If only the TI99-4A had been supplied with 1GB of RAM and TMS99105A CPU then it might have been very successful and beat AMSTRAD to market
Ummmm... If the 4a had launched with 1GB of RAM at the same price point, it would absolutely been very successful, far more successful than Amstrad.
It would _also_ out compete $500k workstations and many mainframes. So, beating Amstrad home computers would be small potatoes. :)
Perhaps you meant 64k? I doubt that would have helped much. If you are looking for "ways to make the ti a market success," the answer is easy: expand the 256 byte "Scratch-Pad RAM" to, say, 16k. That, plus 32k of VDP RAM would produce an absurdly fast machine for the time.
@@jmr5125 Possibly a stupid question, but wasn't the "scratchpad" in-silicon on the CPU? I've always understood... possibly incorrectly... that this was the case. If this was an external chip, on the other hand, well... 256 bytes is what you get addressing byte-by-byte over an 8-bit address (as opposed to, say, addressing by memory words or the like). So, if the "scratchpad" was only addressed by an 8-bit address path, you couldn't improve that 256 byte region without reworking the CPU, anyway, regardless of whether the memory itself was internal or external.
So, does anyone know the inner details of how this scratchpad region was really accessed?
It seems logical, to me, that this would be on-die and would have been intended solely as a temporary holding space for intermediate calculation steps. That we try to use it for so many other purposes really comes down to how they skimped so terribly on REAL directly-addressable CPU memory in the system design.
Had there been that 16-bit (64K max) CPU memory directly connected to the CPU over a dedicated memory bus... that would have been ALMOST as fast as scratchpad, and would have given us the performance that the system was, theoretically, capable of.
But "Scratchpad" isn't really SUPPOSED to be a replacement for "directly addressable CPU memory." Sadly, it was the only such memory we got in the shipped design. And that was a design CHOICE made by someone in a high-level position at TI... probably someone who didn't want to undercut the market for their "mini-frame systems" by shipping similarly-performing units to the home market.
In other words, it reads almost like internal sabotage... (sigh)
"Randomize the entire 64 color palette" ? I thought the system has 16 colors (only 64 with the F18A card)
The VDP is capable of 15 different visible colours (and transparent). But the Standard Mode palette is divided into 32 palette groups (each of which may contain any combination of the available colours). So in memory, the Standard Mode palette is 64 values (4-bits each), each of which signifies a foreground or background colour for one of the 32 palette groups. So when you write those 32 bytes, you're setting 64 discrete values.
@@PixelPedant does an F18A add anything? Is it supported? Like is it simple to work with.
The F18A adds a great deal, but almost none of it can be exploited directly via BASIC. T80XB allowing 80 column mode via BASIC (with an F18A) is one limited example. I have to think someone will integrate a couple other key F18A features into whatever set of BASIC tools at some point, though. Or just as a standalone CALL LINK.
I programmed direct assembler on that machine. Wasn't too hard.
Very impressive but not really XB anymore is it?
Hmmm RXB 2021 has ROLL up,down,left,right and SCROLL too.
All are written in Assembly and will run from only a Console and RXB cart no 32K needed.
Hopefully, RXB will be the next video.