I was thinking...

Why doesn't TI release a list of so-called "supported" ROM entry points which do not change from ROM version to ROM version. Assembly programmers could safely use these; the use of other entry points would be possible but at the risk of being incompatible with future ROMs.

HP has done this since 1990 with the HP48, so I see no reason why TI shouldn't do it.

I also agree!
Companies like TI should support their user base, not hinder it. Allowing programers access to a documented list of Rom calls should be availible to those who want to extend the calc. This could open up possibilties even TI didn't think of.

that the I agree thread was cool... until it screwed up the page layout :)

I agree! :)

I have already returned my HW 2.00 calc to the store and I bought an HP49g. The SDK from HP is excellent. Everything is documented, its almost as easy to write RPL programs as basic, althought the HP49's keys still bother me :)

Traitor...but it was probably a good choice...

. . . which reminds me to dig out my HP49G Emulator . . .

Although I still think the HP49G sucks, I too am going to get an HP49G, just for the support that HP gives it's customers. I still think the 89 is better (hardware) than the 49G, but now in my mind, TI is screwed, and only when (if) they change, I will go back.

-Miles Raymond

You still have a HW 1 calc, right?

Yes, of course TI should release documentation about supported entry points. They exist, and I find it almost dishonest from TI to say in their brochures that their calculator has assembly capabilities if they don't provide any kind of support.

They probably have reasons not to give us the data, of course... Anyone has an idea why they stay silent, and keep their system closed?

They've already sold you the calculator, so they don't care about you anymore. They only care about selling new ones. TI has been historically shody in suporting their calcs. We still don't have all the info on the 86.

And now they want to sell you their $300 SDK.

--BlueCalx

Yeah, really. For three hundred bucks, it better make programs based on my brainwaves.

When was kid I wished that there was a helmet that you could wear on your head(that was hooked up to a computer), think about the game you wanna make(and all of its features) and voila you made a game!

Ghh... True... I'm afraid this shows that TIs sell well enough without real programming abilities.

Probably, too, they think most people would make games if they had better programming tools, and they want their products to look like math-machines, not games-machines.

But I don't think they saw that some people learn programming as well as math, and knowing how to program such "exotic" systems as calculators is always a "plus". They could have made their products not only math-teaching, but also programming teaching.

Maybe so that people can't so easily locate all their symbolic stuff on the calc and reverse-engineer it?

AFAIK they *have* released such a list. It only contains the function _names_ and there is a possability that some functions have chenged places in AMS2 (internally it's a large update), but still - it's right there on their web pages.

Like it says in one of the threads below, don't confuse ROM entry points with the actual locations of the functions. Making updates to a ROM without moving functions is close to impossible. What TI has done with the 89/92+ is put a table at a constant address ($C8) with the addresses to all the ROM functions. This table never changes, so you (or a kernel) can read this table and your program will be compatible with future versions.

Now, if you want to access hardware directly, that's another story. There is no way TI will struggle to keep things compatible at *that* level :P

First of all, I haven't tested the new AMS. Can anyone send me a vti state image (*.sav) of an AMS-2 calc? My address is robertoguerraf@hotmail.com.

I haven't heard of any exciting new features in the new ROM. Installing it is not worth the worry.

Any of these features would make me a lot happier about any new AMS:

*more powerful algebraic processing
*harmonic sum (to get the equivalent resistance of two parallel resistors)
*Ability to stop the AMS from simplifying your expressions
*Small font for long result expressions
*Pretty print in the GraphScreen
*Small font from basic
*Pretty print in small font
*Simplified (stylized) menu titles. I don't like the tab-like top of the menus
*Ability to show/hide menus and status bar so to have more screen space
*Parametric 3D surface plotting
*Symbolic and numeric gradient, curl, divergence, laplacian, etc.
*Numeric fourier transform, compatible with the CBL
*Ability to define operator functions like the Hamiltonian operator
*Integrate the DiffEq program
*Full-log and Semi-Log graphing
*Grayscale graphing
*Ability to do "standard" serial i/o
*Password protection/encryption (the CASIO 9850 has this) to hide contents of programs/text or prevent execution. Programmers could then make a living out of their programs.
*Optimized graphing. I am pretty sure the 68000 can 2D-graph faster than that.
*Ability to set the shut-down time
*Object oriented programming
*On-calc C compiler (with math functions)
*Of course, bug fixes

After typing all this, I think I am copy/pasting it to my letter to Santa :)

I agree whole heartedly! Just one thing: TI will never hear these ideas if somebody doesn't tell them! Also I love my games, but for Caculus the new AMS is worth sacrificing them for!

I like all of your suggestions. I think one of the best suggestions is the ability to stop the calculator from simplifying expressions (like a second-diamond-enter or something like that). It'd also be nice to have the trig functions integrated (like sec, csc, cot) which I know are user definable functions, but it'd be nice to have them integrated.

You gotta be joking about password protection as a means of living for programmers. What are you gonna do, charge people to play your games? That's outrageous!

~ferich

If you want to make money, don't program calculators. Most people program calculators because it's fun (although there probably are a few masochists out there).

If you need a program to do logarithmic scale graphing on a TI-89 or TI-92+, you can download it from my homepage (URL included). If the frame redirection with the "?" in the URL doesn't work, then:
1. Go to http://kevinkofler.cjb.net
2. Click on "TI-89 programs"

You need to download the program "LOGSCALE" from the list. It is in TI-BASIC, so it is compatible with all AMS and language versions. (I have tested it myself.)

I think those would be nice, and one more thing to make it sugar with that nice is showing one step at a time how it got the answer to the problem you entered, like that 49g does, to show the steps to the problem. That'd be something spiffy.

The 49g shows the steps?!? I never knew that. It'd certainly be nice. I mean, right now I use my calc to check my calculus work, but I still have to show my steps. It'd be great if I didn't have to know *anything* and still get everything, including the steps.

In my calculus class we don't have to show the steps on the tests (and homework is not collected), so I just use my 89 and laugh at the poor suckers actually figuring it out.

I'll be laughing at you when your batteries will die :-)

Kaven

Crap, I already upgraded before I read this discussion. Now I am stuck with Basic games. Crap, this sucks.

What are you talking about - you can downgrade. If you can't find someone with a v1.00 or 1.05 then just e-mail TI and request it.

Sweet, I forgot about that. Thanks alot dude.
PS: if you programmers can figure out how to run ASM on ver2.0, I would really appreciate it.

Happy days are here again, no longer have to sit in boring chemistry class and stare at blank BASIC games.

Whooo hooo