I just want to commend Thomas on yet another spectacular release. I love every bit of the new improvments, especially the "train your brain feature". Oh and the menu kick a$$!

Thanx. Sorrily there are 2 minor bugs within this release which were detected too late by my beta testers. Therefore I've to release version 2.01 in a few days.
One bug is a real critical one: Don't archive one of the ticsave files or the ticcfg file! The program won't check until now if a file it tries to write to is archived and this may cause a crash of the calculator (depending on the AMS version).

The other bug is related to the smart take-back feature: the system flags are not correctly set after a take-back, so taking back a rochade and try to replay it again (by moving the king) won't work. As short work-arround: Take back more moves and play them forward again.

*NEWS *NEWS *NEWS *NEWS *NEWS *NEWS *NEWS *NEWS *NEWS *NEWS *NEWS *NEWS *NEWS *NEWS *NEWS *NEWS *NEWS *NEWS *NEWS *NEWS *NEWS *NEWS *NEWS *NEWS *NEWS *NEWS *NEWS

1)HP RELEASES 2 NEW CALCULATORS
2)NEW TEOS 1.02
3)NEW BOMBERMAN CLONE
4)MORE NEWS AT GRAPHITI AND TI-FR

I'm looking for a programmer who may help me out with some programming stuff for TI-Chess 3.0. The important things I need for now are compression/decompression routines which are compatible with the TIGCC environment (inline ASM or pure C - doesn't really matter). The size of both routines shouldn't exceed 2-3kB.
And BTW: If there is someone out who may help in improving the endgame capabilities, please contact me.

This program is great... I've already modified the previous version quite a bit. Added multiple save slots, modified the takeback feature a bit, and a few other things. Guess I'll probably need to redo some of my customizations for this version.

I've played with compression a little bit. If you want to email me some details about what you need then I'll look into it.

-Kevin

The source for GNU gzip is available from ftp.gnu.org, and definately compiles on the GCC compiler. You'll have to basically skin it down to the main compression/decompression parts ( -d is decompression, btw ), and experiment with the block size to get something that will actually run at any decent speed. However, gzip is a good compression algorithm that should do fine at a decent compression level on the 89/92+.

And that's exactly my problem. With a +60 hours job there is almost no time left to do things like that. I've no problem to find source codes of compression algorithms, but the "reduce-to-the-max" step is too time consuming to do it by myself. I'm just glad to finish the 2.00 version one month after the 1.10.

hi i donwloaded tichess and the ggame is cool but after i run it a couple of times my other programs crash. I had a good backup so i ddint lose anything but it is annoying. I ahve rome 2.03 on hardware 1

i tried this on my friends calculators and they crashed too. They all have hardware 2 with rom 1.05


NE way please find out what is worng and somehow make the menu faster. Its gets irrating after a while.

Nothing to do sorry ...
But I was wandering I quite new to TI programming but as I look at all the 3d Engines for TI-89 or Ti-92 something appear there all based on raytracing an Z-Buffer wich is quit expensif in CPU ressources why haven't any of you (TI programmers) tryed some different such as Portals wich I think would enable the engine to be much faster and offer the oportunity to devlop real 3D games... As an exemple I could take Doomling a Fast 3D game runnig on the PalmPilot (TM) and using a portal engine...

Could you explain what a portal is in this context? I do know something about doing 3d rendering, and probably the authors of these games do too.

Remember, though, that the TI-89 lacks a FPU and thus emulates all FP arithmetic, which is quite slow, so the best algorithm will be the one that reduces the number of floating-point emulator calls.

A fast game shouldn't use real FP operations. FP operations can be implemented as fixed point FP where integer values or long values are interpreted as FPs. The real FP operations may be just necessary to fill up some lookup tables and program start.

The only true z-buffered engine I can think of is 3DLib, and that's only just now being used and is quite snappy. ASM based 3D modeling libraries are quite fast, as opposed to a C one. Check out the torus demo for prosit(better yet, run two at a time!), and you'll see these cpus are quite powerful when properly coded for. Raycasting is what you're thinking of, and is actually more complex, but is faster, than true z-buffering. Really, there is no other way to simulate 3D without lots of complexity... and before dashing off to argue, what is Portals?


If images were raytraced, it would take forever for it to appear (a PC can raytrace a 320x240 image in about 2 seconds, depending on complexity, to give you an idea).

I think you are referring to rayCasting, and there are different methods of implementing it, depending on the type of world model you use. The simplest one is occupancy-drid based, like in wolfenstein 3d, where all walls are orthogonal, and so the engine only needs to cast one ray for every pixel-column of the screen, which makes it quite fast compared to other methods. On the other hand, a game like Quake uses a much more complex method based on geometric shapes. It's more realistic but requires a lot more processing power

TI-Chess is great. Other than a text reader, it's the only assembly file I keep on my calculator (out of the constant fear that my HW2 calculator will crash, erasing all the notes I type into it, a problem I've never had with TI-Chess). Great job!

sounds like you need ams 2.03 for the archive-after-crash support (or archive utility if you dont want to upgrade)

This new chess version is great, except are you going to release one that will work under Doors OS explorer? Since that will be great because i don't have to run it using the VarLink menu...
thanks..

go to tinews.net.. they have a working doors version. -Jordan

did you change anything or just add a #include <doorsos.h> at the top?

i just changed the defines.

The DOORS version can be build by yourself by installing the TIGCC environment and just change the compiler define COMPILE_FOR_DOORS from 0 to 1 at the beginning of the tichess.c file.
The problem is that TI-Chess won't run correctly if you haven't installed DOORS 0.96 (problems with relocation of variables in previous versions).
And I don't know how to check from within TI-Chess what version of DOORS is installed.
You see the problem? I hate programs which crashes, but I cannot prevent people from running the DOORS build of TI-Chess on the wrong DOORS version!

I think more games should be written in C like this and Aerox were. It seems to me it is much much more universal (seeing as how it runs on both hw versions and rom versions, or so I'm told).