Yuku free message boards

Well, it's becoming obvious I don't have enough time to cope with TICT software.
Therefore, here is the complete distro of TICT-Explorer pre-1.40. You will be able to download it in a while (there's been a problem uploading it, but I was not completely disconnected from the server, and I have to wait for the timeout) at tict.ticalc.org/downloads...xp.tar.bz2
Not only it contains glitches and lacks a number of minor things (see the file todobugs.txt in the src directory) that have to be fixed, but most of all, the launcher part remains incompatible with the TI-89T. None of the suggestions made above by Kevin were implemented. If nobody does the job for me, it's pretty simple: TICT-Explorer will remain as is (no Titanium-compatible version) forever.
Of course, I need real help - not "help" that will force me do the work again, since this is simply impossible for me. By that, I mean that starting from now, like all other TICT software launchers will be, tictex has to be available in at least four versions (name mapped onto ttstart): tictex, tictex-titanium, tictex-universal, tictex-slow. Fast decompression routine for the three first ones, slow decompression routine for the last one (for this one, please wait for the next official version of ttstart).
In the future, I plan on making tictex-sstart, in order to save space in case both TICT-Explorer and SuperStart are used on the same calculator.
TICT-Explorer 1.40 must be on-calc incompatible, just as TICT-Explorer 1.30 is.

To build TICT-Explorer, use the.tpr files (tprbulider) and then build89 and/or build92p.
Copie de tictex.89z is the old launcher (TICT-Explorer 1.30), which I usually use to test the tictexpl part.

> And I find it ridiculous how on one hand you complain about me supposedly imposing my personal opinion on TIGCC users when you are trying to impose even more of your own opinion on future TICTEX maintainers.
YOU are ridiculous:
* you do impose your own opinion on TIGCC users, it's a documented FACT.
* imposing your own opinion in TIGCC potentially screws up all users, imposing my opinion on future TICTEX maintainers does not screw up anyone.
* your count of the different TICT-Explorer versions is nonsense. I said which are the versions: tictex ("fast" implied), tictex-titanium ("fast" implied), tictex-universal ("fast" implied), tictex-slow (tictex-universal with slow decompression routine). Exactly the same as ttstart. There will (may) be a tictex-sstart later.

> Guess what, I'm going to ignore it anyway, either I'll simply release a forked TICTEX ignoring your requirements
Please do ! This way, you'll miss all the optimizations and bugfixes I did, which decrease the size of the whole TICT-Explorer by 6 KB. Then you're going to explain us that everything should be optimized for size and that you miss the time where TICT used to be symbol of bug-free programming...

I received no report about the tictexpl part of the program (which you can drop into an existing TICT-Explorer 1.30 installation, I did not change the magic numbers on purpose). Remember that it is ~6 KB smaller, has more features (although none of those mentioned in the first post was implemented).

As you may have guessed since this topic was written, with the release of ttstart 2.00, there are no more on-calc incompatible versions of the tictex part.

I had already found that bug a while ago, running it on a TIEmu without HW3Patch. This is one of the things I properly tested, although I missed other severe bugs.
Given that the fix was so trivial (a non-HW3Patched 89T cannot do much), I let the source in that state. I didn't think about detecting whether HW3Patch was installed, and do nothing in that case. Thanks for the fix.

[EDIT: as it is, the fix would turn into large code due to HW_VERSION. I've been using something else at other places, like
if (!(TI89 && ROM_base == (void *)0x800000)) {
in tictex.c, instead of TI89_TITANIUM which is (TI89 && HW_VERSION>=3).]

Hey, I'm not the one who invented the following hack in SIDE, on which TICT-Explorer 1.40 is based:
unsigned long offset = (~*(long*)0xc8 & 0x800000) >> 5;
Are you ? Likely.

Anyway, I bet that our friends at TI will break something more in the next model, as they always did in the past. All programs would then have to be updated anyway...