BrandonW has just released a patch to get xLIB commands to work on OS 2.54MP, TI-84 Plus OS version that the TI-Nspire OS 2.0 and 2.0.1 emulates. Until now, those commands would do nothing, even when xLIB was enabled. It should come very handy for people who want to run games and programs that won't run fine under Celtic III.

It was added to our archives here.

That's good! Now I don't have to worry so much when I make games. Speaking of which, one of these days I'll get off my lazy rear and get back to work on LNBL and my other games `-`

Again, BrandonW's proven how much he can do to fix TI's errors.  Thanks Brandon!
Do you mind telling us what you had to change?

It's on brandonw.net. They basically re-arranged a lot of code in the parser and broke what happens when canceling from the parser hook (which is what xLIB does when it handles a real( token). They set numOP1,(iy+ParsFlags2) when they really shouldn't be, which confuses the parser because it's looking for a result in Ans/OP1 when xLIB didn't return one (or returned garbage).

It's highly dependent on how parser hook-using programs are coded -- this is why Celtic III (coincidentally) doesn't have the problem.

I sent an e-mail to TI explaining the bug and how to fix it, but they'll probably either ignore it or it won't make it to someone who can actually fix it.

Yay! I tested it on Age of Culture II and it worked great! Thanks a lot!

And also, does this also fix the Omnicalc problems? I didn't dare try it on my calc...

I don't know specifically what the Omnicalc problems are, but I doubt it.