More people should make more games that can only be played with the Expander2 (because of the memory they would take up). This would let these games be bigger and a lot more graphical. Games like the FF series maybe...

I highly disagree. There would be so many people not able to use the games, that I doubt that authors would even bother. 25kb on the 85 is kind of meager for certain things, but 64kb on the 86 should be plenty for most anything.

That's a pretty good point, Kirk, but having an E2 option to pull all the large map and graphics info off external memory would be very cool for certain stuff. While I obviously think everything you said applies to E2-only games, it would not be very difficult to have the game probe for the E2 on startup and then use smaller maps/lower quality graphics if one isn't found. The TI-85 memory is so small it precludes having any large Sorbo's Quest-style games... you almost need something like the E2 for programs with really large amounts of binary content on the 82, 83 and 85. The TI-92's memory also fills up quickly due to the larger opcodes. As far as the TI-86, 89, and 92+ are concerned, there is clearly less need, but it would still be pretty cool--especially since you can design an IR linking module for the E2 and have both larger maps and some pretty advanced multiplayer. This stuff is theoretical now, but it could be done with not too much effort by someone with your talent.

Of course, I may be a wee little tiny bit biased having designed the whole bloody thing. ;-)

-Bryan

Try compression instead for large maps and graphics. It worked on the NES. Just look at SMB1, world 1-1 is 101 bytes...

And the expander only works with 85/86/...

Well, the code to use the E2 has been readily
availible for some time. I encourage programmers
writing games to consider adding support for it,
since all the tough routines are already written.
If you're interested in writing a game with the
E2, feel free to contact me and I'll give you
whatever help I can. Maybe even an E2 parts
discount. ;-)

The biggest reason more games aren't using the E2
to its max potential is that there are relatively
few E2s in existance -- only around 300 now. But we're going to be shipping that number over again in the short future, so hopefully more stuff supporting the E2 will be coming out soon.

---
Bryan Rittmeyer
mailto:bryanr@pair.com
http://horizon.pair.com/

A better solution would be to have something similar to the TI-89's archive capability, but built into a shell. Like for the 86, zap-os could be like plusshell and automatically unarchive stuff. The 85 is a different story, because most of those shells are dead :)

I believe the biggest fall-back of the E2 is that it is useless w/o it's driver... What good does it do you if you have all the 85 games in the world on your E2 but your 85 just crashed at school? A lot of people don't have 85's at my school, so you probably couldn't find anyone to give you a shell, and they probably wouldn't have the driver either...

I don't think the E2 is bad, far from it, I want an E2 very badly =) but I think that it could use a few improvements, like a button on the outside of the E2 that sends the driver(s) to the calc...

-Miles Raymond

I agree that sending the driver would be very nice... in the late developmental stage I took a prototype on vacation and sqrxz locked my TI-92 hard and I lost the driver... That was not a fun way to start vacation. ;-\

However, even though the E2 has a lot of free horsepower, sending the driver is just not technically feasible without a lot more hardware. The E2 speaks only the lowest level TI protocol and can't send variables using the ROM routines. Trying to make it do so would be a huge task with very little benefit. I mean, how often are you completely away from a PC for an extended period of time? I know having to sit through class without games is tough =) but you can always re-send the driver when you get home. Crashes are annoying with E2 or without; if a program you like is crashing a lot, you may want to contact the author and yell at him/her. Sometimes we need the encouragement.

--Bryan

Or do what I do: simply not play the game at all. If I load a program on my calculator and it crashes repeatedly at school while I play it, I simply don't load it back onto my calculator again. This helps to almost completely eliminate potential crashes on my calculator.

--blue

for the 89 user, use au 2.00 and you will never need it again

au 2.00? wassat ? have I missed something :)

I see no mention of 89 in the message you replied to.

From what I gather, nothing can get on to the E2 without first being on the calc. If there was a program larger than the calc's capacity, there would be no way to get it onto the E2.

I think that you should make an expander for the 82, 83, and 73, when you are finished with the others.

For the 73? Is this really necessary, considering the incredible lack of programs for this calc?

The expander2 seems really cool. I put my name on the list to get one last August....and I still don't have one... Oh well.

Maybe this is not the best place or time to post this, but I have an idea which may prove interesting to all of you.
Note: Before you read on -- this is entirely off topic!!

With the high level of Gameboy-like games coming out from the awesome programmers, I was wondering if it would be possible to take it one step further. Many of you know that Gameboy emulators are widely available on the Web. Some have their source code included. Now, someone with incredible programming ability (not myself) could port one of these to an upper-level TI calculator (89/92+) with a little bit of work. I have seen DOS versions of these emulators as small as 70K. With a little tweaking, that could probably be lowered, especially if you drop sound support. Now, here's where the problem comes in, the games. GB carts were made in varying sizes. I know that Zelda, for instance, is 512K, though MarioLand is 64K. Zelda, of course, is a bit too large to work with, but we have the potential of lower-level games. Lots of good Gameboy games are 128KB!! (But don't even consider games like Pokemon - they're 1MB+)
Now, this would be difficult - though, I think possible - to pull off on several fronts:
1) Screen size
2) Memory allocation
3) Complexity
4) Compatibility (how to store the game on the calc?)
5) And many more... (fill yours in here)

This is just a wishful post by a guy too busy to learn ASM on his own time, but those of you who are able to -- what do you think?

Tony

P.S. Flame, critique, comment all you want -- It will keep you busy and I don't care what you think of me :-) Also: If you can't do it now, I will - after 4 years of college and programming classes, ok?

I actually suggested something like this on Macross' message board a while back, but nobody seemed interested. While I don't personally own a TI-89 or '92, I do have some ideas that may help...

PORTING- Since these calcs are based on the Motorola 68000 CPU, it would probably be easier (though still very time-consuming) to port a GameBoy emu from Macintosh. That way, the emulation core doesn't have to be written from scratch.

COMPRESSION- There are various ways (though I'm not familiar with them) of creating compressed programs. This would be very useful in the case of an emulator. Compression would also be handy in the conversion of the ROM to a TI-compatible file format (ROM images are notorious for filling up one's system :]).

LINK PLAY- One cool thing to add (it's only supported in one Gameboy emu, NoCash GameBoy, because it just really isn't feasable on a computer) would be GameLink emulation. Then you could link-play the REAL Tetris (or Mortal Kombat, or WildSnake, or whatever) against a friend in class.

Well -- how do we go here?

According to Kirk, something such as I have mentioned would be so worthless to even attempt that no one would.

Something else I was wonderering... The Gameboy has to have some low-level ROM built in to hadle certain functions, doesn't it? That would be all we need to emulate, wouldn't it?

I don't know. This is going to come to wishful thinking, but it seemed possible to attempt at least. However, I put trust in those like Kirk who know the abilities of the MC68K.

Maybe we should table the idea until TI introduces a more powerful calculator... what do you think?

Tony

P.S. Sorry if I offended you, Kirk, in proposing an "impossible" idea. Consider it something to "change the topic" of what discussions have been centered on lately.

Three things need to be emulated to be successful. First of all, the actual processors are different. Thus an emulator core is needed to execute the instructions (this is what will make it run quite slow). Secondly, the hardware must be emulated (what fun, sound, sprite capabilities and all...). And finally, the screen must be emulated, which is a different size than a calculator screen. Take all these together and put it on a 10mhz processor and... well... have fun :)

I think that the following machines are worth emulating on TI89

ZX Spectrum, this emulator is almost finished, Z80 3.5 MHz machine, with a lot of software. Amstrad gave copyright permission for use of ROM.

Apple II+, this model was quite simple because it had no page swapping, processor is 6502. I do not know about ROM copyright.

ZX81, Simpler than ZXSpectrum and as it generates picture in software with slower processor clock, we can reach 100% speed. Amstrad gave copyright permission for use of ROM.

CP/M, software is not much clock dependant, and there are many compilers and WordStar. Source is freely available from Caldera.

Any stronger machine will not be possible to emulate on 10 MHz 68000 processor, but even these four machines had more software than all TI, HP, Sharp and Casio calculators together.

Umm... Let's get something straight. The Gameboy uses a Z80*like* processor. But it's not Z80. That means that even a Z80 calculator would not be able to emulate the gameboy without parsing opcodes, plus all of the hardware stuff. But M68000 - that is a totally different processor! It would be terribly slow. If you were to run a gameboy emulator on an old 8mhz pc (assuming that you could find one that would even have graphics), you would see that it would run insanely slow and have no real purpose.

And finally... It is a calculator, we're lucky enough to have all the games that we do have. If you want super high quality games... buy a Color Gameboy. Save the calculator for the games that can be played during class.

Hey, here's another idea:

Attach a calculator to the back of a Gameboy! Wait...the circular bulge on the Color Gameboy would make that too hard. Plus, you'd have to remove it to change batteries for either system. Oh, well...a two for one deal anyway!

Not exactly about Gameboy, but ...
Currently I write the emulator of ZX Spectrum for TI89, in fact I am in final phase of bug fixing. Many hits already work: Hobbit, Atic Atac, 3D Ant Atac, Total eclipse, Fred, Knight Lore, Manic miner, Jet set willie, Elite, Hungry Horace, Lords of Midnight, Footbal manager, while some other games still do not (Chuckie egg, Super Hang on, Penetrator ...). The speed of emulation is 30-80% of original ZX Spectrum (Z80 at 3.5 MHz, 16K ROM, 48K RAM).
The emulator requires about 140 K of user memory and at least 19K of archive memory.
ZX Spectrum games were miracle of the optimal programming. Elite put in 48K the world of 2500 solar systems and 20 types of ships. Lords of midnight was adventure of 32000 location pictures. Cyrus chess in 16 K won many championships. Driller was DOOM like 3D game with shaded objects, written 5 years before DOOM. Jet Set Willie was the most played game, and only a few successed to finish it without pokes for infinite number of lives.



On www.void.demon.nl you can find about 7000 Spectrum programs! I belive that more than half of it will be possible to play on TI89 on my upcoming emulator!

Samir Ribic
Sarajevo
Bosnia and Herzegovina

Nothing like squashing an idea, huh Kirk? ;)

Maybe it_is_just wishful thinking, but Tony didn't seem to be doing research for reasons on why his idea wouldn't work. He was just casting an idea out and seeing if anyone would snatch it up.

BTW - no one ask me what_my_idea on how to proceed would be - I have no ASM experience and willingly admit it. :)

Thanks for your support; I'm glad we're in this together. (As I said, I have no ASM experience).
You are right - I was merely proposing the idea to gather others' opinions, especially that of programmers.
I don't hold anything against Kirk because he comes from the viewpoint of an ASM programmer, and he knows the limitations of the calculator.

For now - I will be content using it for what it is and for the games that are out already. I hope others on this site take that to heart before they burst out in rage to comments (especially on the other message board pages).... :)

Thanks, Tony