I like this idea more for actually using the programs. If I want to group all my programs in folders and stuff, DCS is my peice. I hope this gets out there.

And will it have compatibility with compressed programs that only thing like CrunchyOS can run? With that, this will be the ultimate gamers shell.

What about archived programs? or do you have to unarchive them? *looks through thread* Yeah, it will run archived...

Kk, I'll look for it teh first time it is out to the public on something like TI CALC... Cool!

Yeah, it's kinda in need of debugging right now...

Yeah, I kinda have known that from the previous posts

But I was saying when it comes out perfected I will definitely use it.

Edit: 200 Posts!

Same here.

Me too!

Re-uploaded. It ought to work better now.

Hmm... you mean I can execute ION and MirageOS program from BASIC programs now? Hmm... [evil grin]

Yes, you can.

Evil and excited grin it is (after debugging)!

I would prefer to NOT have a GUI. On a calculator they are clunky, and don't work well with the given HID input methods. This noshell sounds awesome - once the major bugs get worked out, I will certainly give it a try. Just curious though, I hear that executing an ASM program in a Basic program is slow (I mean the "regular" asm(prgmSOMETHING) method) - how does this compare? Also, not sure how easy this would be to implement, but if you could make it so programs can take arguments that would be sweet...

Looks like a cool "Shell". I should try it out sometime... Although I don't like having my Ram Reset

Guys, I fixed it...really. Try it.



I don't see how using Asm( is "slow", but what it does do is insert memory to userMem instead of relocating the program there (which all modern shells, including Noshell, do). This has the advantage of not needing any extra RAM to run it, and program writeback requires no extra effort.

What do you mean by programs taking "arguments"?

You know, arguments - like for function calls? Say, prgmSOMETHING(arg1, arg2, arg3). I have no idea how practical this would be to implement, or how the program that is taking the arguments would receive them - I was just throwing that out there.

I don't like to say things are impossible, but that would be fairly hard. Actually, I don't want to say hard, but a lot of parser ickiness I don't think anyone wants to deal with. You'd have to manually read in the left parenthesis and the arguments and store them somewhere. Would these arguments be passed to BASIC or assembly programs? Where would be a good place to store them? Lots of questions.

But not impossible.

EDIT: Also, about the above couple of posts, Noshell relocates Ion and MirageOS programs to userMem, but it COPIES nostub ones (because this is how they usually work). I chose to do this because for relocation, it's important that you regain control so you can clean things up. That ISN'T something you can guarantee for nostub programs (some look themselves up, delete themselves, do something, and jump to _JForceCmdNoChar) and that would be very bad if I relocated it. Ion and MirageOS programs, however, were designed from the start to return back to the shell, and so I can relocate them.

Not that any of that is really important in any way whatsoever, but there it is.

I finally got a chance to try the shell out, and I believe it's awesome. I only had one problem, xLib + noshell = not good. When I had xLib enabled and had noshell chain install, whenever I ran Hunters (game that uses xLib) it ran slow, but it worked. When I tried running an assembly program, for example Falldown, all of the graphics were messed up. I tried Phantom Star too and that didn't work either. When I enabled noshell without chaining it, all of my games worked, except for Hunter, which crashed in the middle of running it..

Sounds like Brandon's not using a failsafe fastCopy routine?

I doubt that's it. The hook chaining is a mess (so that explains all that xLIB/Hunters stuff). I'll look into it.

In the meantime...don't chain.

ok. Hopefully it won't be too hard to fix. Good luck.

I had a thought, I belive Kerm finishe runprog for DCS so couldn't add somthing in so that if the program is in DCS asm you use that to run it. But first you would have to cheack to see if DCS is even on the calc.

Hmm, maybe chaining was causing all of my problems...