Omnimaga
Calculator Community => TI Calculators => General Calculator Help => Topic started by: whitevalkery on March 06, 2010, 11:02:10 pm
-
also i have another question... i just tried something that uses xLib on my TI-84 and for some reason no mattyer what i do to enable the xLib any applications that require it just simply say it isn't activated... or basic programs simply crash... (such as this program: http://www.ticalc.org/archives/files/fileinfo/387/38746.html ) (and this one: http://www.ticalc.org/archives/files/fileinfo/396/39695.html )
is this caused by the new OS?
-
Most likely. Try using Celtic III instead.
-
ok
i am testing that right now
edit:
it seems to help... but the programs are all messed up...
this could be either caused by not using xlib and using celtic III instead, or the new OS again...
since now some of the programs run, but have their screens and controls messed up...
-
yeah the compatibily issues are annoying sometimes. Some games like XXR and Metroid II will run not too bad, but others will mess up very badly.
One issue is that certain xLIB games abuses xLIB bugs and glitches to perform some special effects. These glitches are not all emulated in Celtic III.
The only solution is to downgrade to OS 2.43, which you have to google or download from french TI sites such as TI-BANK
Even Omnicalc messes up sometimes in 2.53MP. I also had issues with anything using reverse text (white fonts on black)
I even heard stories about some new OS functions being buggy, even wehn you have absolutely no ASM program installed x.x
-
wow...
thats just stupid...
why would TI release such a half ass new OS with sooo many issues?
-
They most likely released it to combat All the third party OS's in development.
-
but they aren't even out yet...
i have not even seen or heard of one btw... :-X
could you show me some by links?
or is someone on this site already developing one?
oh man... i am going to start looking as soon as i finish writing this post
-
There aren't any finished, but TI knows the community is up for something. Don't you find ironic that a new Nspire OS came out the week after the exploit was released?
I think people such as Brandon Wilson are working on new OSes, but I think the community looks more forward for trying to patches the issues in the ones TI release, since this would be much less work.
As for bugs, this is not new. TI calcs had bugs since the day the TI-81 came out. On the 83 and 83+ (early OSes) it was possible to crash the calc through an Equ>String bug. On 83+ OS 1.13, by scrolling the prgm menu with 50+ programs, the RAM cleared. I heard stories about Nspire OS bugs causing reboots. Finally, the classic memory leak in TI-BASIC caused by using a Goto inside a loop/condition block.
There is a topic listing glitches for TI-83+ and 84+ OSes from 1.12 to 2.43 at http://ourl.ca/3687
-
wow.... this is really really interesting...
I AM SOOO GOING TO ABUSE THIS LOL!!! :o
thx for posting that! my friends will be surprised next lol :P
btw... i always play harmless jokes on their calcs... plus they never put anything in their ram so reseting doesn't cause them any inconvenience...
and they regularly reset their calcs anyways... :)
thx for posting that! now i can write and rewrite some of my old prank programs lol
-
ah ok well try to not cause anything bad XD
The farthest I went is decreasing their contrast or installing fake viruses on their calc. Just make sure they have nothing important on their calc first, and I recommend against the group glitch and the equ>str one. They are very dangerous and could damage the calc.
-
ahh... i see... i will take that in mind...
i dont want to break their calcs lol...
the funniest one i did was the infinite garbage collection program lol...
it was the first assembly program i ever wrote... it basically shows the garbage collecting screen for ever lol...
one of my friends waited an entire period at school and then ran to me asking wtf did i do to it lol
i simply pressed a key combination and it got out :P
oh... and he didn't just pull out the batteries because he knew that during a garbage collection if you interrupt it, it would seriously screw up the memory...
(i found that out the hard way :O)
-
WAIT! I wish I had a prog like this back when I coded The Reign Of Legends series. In these games, we had poor asm libs compared to xLIB and Celtic III, and we had to archive/unarchive a lot to fit the game on calc. However, this caused GC's, and sometimes we smashed enter too fast, and it exited the game with ERR:ARCHIVE FULL. Didn't save progress and got pissed. A program to force garbage collections every few battle would have been very handy.
However, today it wouldn't be as useful anymore to many people since the old method of archiving/unarchiving has been obsoleted by xLIB prgm copy functions (copying programs from archive to temporary RAM files)
Anyway back on topic, I hope BrandonW or someone will release a patch to fix all the compatibility issues in 2.53 MP. Mapar007 did a patch for 2.43 to disable the teacher feature to lock down the PRGM menu on 84+ calcs.
-
The prog he talks about simply shows the screen with 'Garbage collecting...'. It doesn't actually do anything, if I understand it correctly.
To force a GC (that means: no yes/no prompt), you need to call an OS-dependent address in memory. I've got them for some versions, but as I don't have the binaries for all of them, I can't work them out for every existing OS. There is no BCALL for immediate Garbage Collections.
-
aaah ok I see. Thanks for the info
Btw, do you know if it's possible to change the garbage collect message or show none at all? For example, in a game like ROL3, I have a menu option to garbage collect manually, but it would have been cool for example if it automatically did it then right before I launch the command, I display "Reorganizing Archive Memory, this may take up to 1 minutes". Newbies, including myself back in 2001, can be scared when seeing the name Garbage Collect. By seeing Garbage, they may think if they choose yes, that their memory will be cleared. In french, TI fixed this by translating Garbage Collect to Reorganize Mem.
-
This is the worst thing about Garbage Collects. Whenever someone sees "Garbage Collect" for the first time, they panic. I sure did. I deleted programs instead of Garbage Collecting them. :D
FFTOM2's readme has one of my favorite descriptions of Garbage Collects. It compares it to a library, and makes the user no longer scared of Garbage Collecting. :)
-
Btw, do you know if it's possible to change the garbage collect message or show none at all?
Yes and no. This is a consequence of the stuff I explained above. I'll elaborate:
When the OS wants you to do a GC, you get the message prompt, and you select yes or no. Then, the OS calls _PutS (the equivalent of BASIC Output() ), to display 'Garbage Collecting...', and subsequently calls the routine I mentioned in my previous post. So, again, I could write a prog that does this for you on OS 2.43, but it won't work on any other version.
Theoretically you could change the message using some clever flash wizardry (using the extra RAM to backup entire sectors of flash, modifying them, erasing the original, and then copying them back), but all my experiments have failed so far, mainly due to emulators not letting me see ALL physical memory at once, which isn't great for debugging... (asm 1337 and hardware gurus, please correct me if the theory is flawed)
Or, you might just be able to hook the bcall, but I doubt it. Lemme check that if I've got the time. Maybe Brandon or Ben knows.
I hope this is clear enough. Like you, I'm not an English native, so, well...
-
Why did TI change the GC routine in every OS? :O
-
They didn't change it, it just ended up somewhere else (moved by, say, 100 bytes) due to other changes.
Imagine this to be the structure of OS 1, where CCCCC is gc code.
IIIIIIIIIIIIIIIICCCCCCCIIIII
TI now changes some stuff:
IIIIIIIIMMMMMIIICCCCCCCIIIII
The GC code is now 'pushed' forward in memory.
-
aaah ok, so there's no way to directly access it automatically? :(
I thought it could be since from OS 1.12 to 2.53 MP, several Ion and MirageOs games continued to work perfectly or with very minimal crashes. Celtic itself works from 1.13 to 2.54. I never tried on 1.12, though.
-
Wow, that's a lot of OSes to work properly on. Props to Iambian! ;D
-
I FOUND IT!
List of all past (and present?) TI-83+ OSes worked on in the TI community!
http://wikiti.brandonw.net/index.php?title=83Plus:OS:TIOS_Alternatives
I have never tried any of them, though.
By the way, since many of you are new, did you know that there was a time where people would work on third party OSes for the 83+ and 89, then Texas Instruments took action against the developers? There used to be a time where TI wanted to disallow third-party OSes development for their calcs. I don't remember how, but the community convinced them to let developers do such unnoficial OSes. Finally, all that happened is that installing such OS on your calc would apparently void the warranty.
I am currently looking through old Detacheds Solution topics for more info, same for UTI, because it has been so incredibly long.
And in 2002, here's what TI did regarding unlocking flash?
http://www.detachedsolutions.com/forum/mv/msg/240/0//
EDIT: Thanks to Jonimus/TheStorm in #ti:
[20:19:46] <Jonimus> DJ_Omni: they basically said we can develop the OS but not distribute the methods of installing them
However, this changed this summer, with the key factoring and the EFF fighting the DMCA notices.