Wow! Just last night I was pondering why there wasn't an 89ti PreOS! Thanx Patrick! PreOS is my favorite shell, it just works so easily. I guess that is why it was one of the first programs I reviewed :) LOL.

There already was, it's called Iceberg.

Well yeah, I was one of the ones who contributed to the article about Iceberg. The problem is that Iceberg is not as fullproof as this new PreOS.

How do you mean that? "Foolproof" as in "automatically patches kernel-based programs"? That is pretty pointless considering it won't do anything about _nostub programs, so you need to use GhostBuster anyway if you aren't sure that the program needs a kernel.

PreOS tells you if it needs a kernel.

Perhaps the news of your works, as well as others, needs to be spread around sooner. Iceberg has been available for a while, hasn't it? (At least the beta?)

> Version 0.70 is fully compatible with the TI-89 Titanium

Iceberg 1.00 is, too.

> and runs on all AMS versions and all hardware versions.

That's nothing really new either, all run either PreOs 0.67 or Iceberg 1.00.

> The grayscale library now works on all hardware versions as well.

Nothing new at all. HW3 grayscale is just like HW2. And PreOs 0.70 doesn't even have the TitaniK backwards compatibility modifications to the grayscale routine that Iceberg has.

> All kernel programs should now run on the Titanium without patching.

I'm not so sure, the autopatcher in PreOs 0.70 fixes much less things than GhostBuster does. And I also don't see why it makes sense for PpHd to run the patches at each execution. IMHO, GhostBuster is a much better tool for the job.

You have obviously failed to realize that Iceberg is not fullproof. It actually has a bug, that totally disables the STDLIB interface at times.

That's not Iceberg's fault, it's you who are using the wrong libraries! Stdlib are PreOs libraries. There is no such thing as a stdlib in Iceberg. (You can build one yourself from the Iceberg libraries if you want though.) Some PreOs 0.67 libraries (such as genlib) are compatible with the Titanium, but others need to be patched with GhostBuster, and for graphlib and userlib, you definitely should be using the Iceberg version!

You still can run "PreOS Browser" from Iceberg. Send "shell" from Preos 0.67 to your calculator, and run it. You can also patch Iceberg so that it runs "shell" instead of "TI-CHESS Explorer".

I don't whether the programs which are patched by ghostbuster can run on HW1 & HW2, but PpHd's patch doesn't modify the archived program, so it can be interesting

Indeed, in general, GhostBuster-patched programs can't be run on non-Titanium due to the use of the new ghost space for interrupts.

But nothing keeps you from keeping the original version around to send to other calculators, the Titanium has plenty of archive memory.

Not so much if you kept the given Flash Application (~900Kb of Free Archive Memory).

Hot damn! There is a lot of stuff that is not needed that comes with the calc, though. I cannot remember which ones, but when I saw the list I could only remember wanting to keep like one or two. One can send a flash app bact to the comp, can one not?

-W-

>>One can send a flash app bact to the comp, can one not?

No. However, if you want to delete it temporarily, delete it from the calculator and keep the free copy that you can get off the internet on the computer so that you can send it to the calc at the time of your choosing.

Actually, if you have a recent version of your linking software, you can. TI-Connect definitely can backup FlashApps for you.

That is sweet. You never know if you might want one of the paid ones in the future.

-W-

If you have the calculator, just type in your ID number on the website to get it. If you had to buy it separately, just don't delete your computer copy.

Does anybody know if TiLP does app backups?

1.5?

So, you are telling me that I should use iceberg over PreOS?

-W-

Since Iceberg is a port of PreOs 0.67 _HE_ (Kevin) made, yes, he tells you that.

Iceberg is a port a Preos 0.67 which now outdated. There are only 2 known programs which are for the titanik kernel, so using preos 0.70 is perhaps better

Okay, Austin is confused. I am going to wait for further developments.

-W-

If a program requires a kernel, use PreOS 0.70.
If the program doesn't work in PreOS 0.70, run it under Iceberg and see what happens.

If a program runs by itself, use Ghostbuster.

If a program is in BASIC, find out if it calls anything that is patched. Add a patch and then run it normally.

It shouldn't be this complicated. Until there is a TSR or shell or debugger to do this (probably Morvlon), you will have to do it yourself.

Moreover PreOS 0.70 can run theses two programs (Just download them from ticalc archives).

Iceberg = "Titanium Preos v0.67"

More than 97% of the sources are identical between PreOS 0.67 and Iceberg v1.00.

There's still all this. ;-)

History:
TitaniK 0.10 Kevin Kofler 2004-02-14
* First release. Forked from PreOs 0.67. All files modified.
TitaniK 0.11 Kevin Kofler 2004-04-07
* Changed version number identifier from $FF00 to $0011 (PreOs.asm).
* Updated graphlib and userlib with a hack to make interrupt handlers work.
TitaniK 0.12 Kevin Kofler 2004-04-11
* Updated version number identifier (PreOs.asm).
* Using ROM_CALL instead of hack to get tios::Heap (PreOs.asm). Using another
hack to put it into an address which fits in a word (!"§$%&/()=? SMQ ...).
* Removed the error frame saving/restoring hack (Preos.asm, exec.asm). Using
ER_success instead.

Iceberg 0.90 Kevin Kofler 2004-06-21
* Merged Titanium-compatible code from TitaniK 0.12 with TSR code from PreOs
0.67. Readded a few hacks that are needed for crash protection and should
really work. Reenabled the error frame hack because it is needed for crash
protection and because I fixed it for AMS 3.
Iceberg 0.91 Kevin Kofler 2004-06-24
* Updated version numbers (PreOs.asm).
* Fixed retval hack in exec.asm for AMS 3 (the hack in TIGCC works by accident,
the one in PreOs/Iceberg had to be fixed).
* Added uninstaller (forked from PreOs 0.67 uninstaller).
Iceberg 0.92 Kevin Kofler 2004-06-27
* Updated version numbers (PreOs.asm).
* Reverted patch to support event hook convention 2.00 ('EvHk' signature) applied
by mistake, it is not compatible with the TI-89 Titanium (only event hook
convention 3.00 is, and it only uses the original 'evHk' signature) and not
supported by PreOs 0.67 either (saves 24 bytes).
Iceberg 0.93 Kevin Kofler 2004-07-04
* Updated version numbers (PreOs.asm).
* Now saving and restoring the screen rather than redrawing it, so ExePacked
kernel-based programs can now be run using Iceberg and ttstart-titanium.
Iceberg 0.94 Kevin Kofler 2004-07-14
* Updated version numbers (PreOs.asm).
* Fixed the detection of the 2 AMS routines used by the nostub crash protection.
Iceberg 0.95 Kevin Kofler 2004-07-26
* Updated version numbers (PreOs.asm).
* Added ugly workaround to get programs using the old definitions of
tios::font_small and tios::font_large, which are no longer valid on the TI-89
Titanium, to display correctly. (Programs doing custom arithmetic to get those
offsets, such as Solar Striker, still need to be fixed by hand though.)
Iceberg 1.00 Kevin Kofler 2004-08-11
* Updated version numbers (PreOs.asm).
* Don't leak 4 bytes of stack in _nostub anticrash (Ints.asm).

Honestly, did you really need to say that you updated the version number in the source on the updates? I mean, people expect that to happen.

-W-

He might have used a command that affects compilation if the version number is wrong.

The autopatcher fixs what is needed for Kernel Programs.

Kernel-based programs can contain the exact same mess as _nostub programs. There is nothing whatsoever guaranteeing that kernel-based programs won't use some ghost base address other than 0x40000, 0x40064 or 0x40074, or try to compute ROM_base through an AND with 0x600000 or to detect the calculator in that way. And indeed, kernel-based programs compiled with TIGCC are likely to contain some of these. For example, the grayscale routine always uses its builtin hardware detection, which had the ROM_base bug before I fixed it for the Titanium. Programs might also be redirecting AI2, AI4 or AI6, not just AI1 and AI5. Your patcher is very incomplete.

Hoppefully, Kernel Programs are now much cleaner than Nostub ones!

Err, build up on this. I'm curious to learn what you mean.
On the contrary, there are kernel-based programs that use so nasty and unsafe hacks that they won't work on anything else than HW1 AMS 1.00/1.01/maybe 1.05 without modifications (which were not necessarily done)...
Those hacks were supposed to give more optimized code, I guess ? This is exactly the bad use of optimization, on-calc incompatibility is not.

It seems you lived in the past.