Even with some improvements, TI's proprietary Lua cannot - all the more it's not JITed - match the speed and power of native code, that Kerm is enjoying without restrictions on the Prizm

Indeed, Lua will never match the speed of native code, but that doesn't mean it's too slow

It also depends on the skiils of said programmer, I can make pretty good scrolling games with Lua, just because I have experience with the platform.

Depends in what context you mean. You may be able to do some simple to slightly-advanced games with interpreted Lua on the Nspire, but the potential that the prizm has is still yet to be seen with its native development. I think we're already seeing the most advanced stuff that will come out of the Nspire's Lua; I however, think that programs on the Prizm will be ever evolving for some time to come.

Well, I like the gameplay concerning the way when you complete the goal (the other player hitting the ceiling, a time limit, or a line limit), it gives you a token with a letter on it. When you get three, they spell "WIN" and you have won. Thus, you play three rounds. I also like the way it can clear all the lines at once, I.E., when you get four lines cleared at the same time, it clears them from left to right, as opposed to Tetris' like Bill Nagel's, where it goes clear, clear, clear, clear! One more: you get higher points the more lines you clear at one time in the SuperNES version. Hope that is more specific.


P.S. Can you add the BGM?

I'm not saying the prizm is bad :p (I would gladly program on it if I would have one)


Oh no, just wait and see

Caleb: What's the BGM? And I like the idea of clearing sets of four lines side-to-side; currently Tetrizm flashes the set of four lines together, then removes them at once. And I already have you getting more points for clearing more lines at once.

Jimbauwens: I think I see Ashbad's point, that that best TI-BASIC programmers are now getting to apply their skills to TI-Lua, trying to get a somewhat limited JIT language to do cool and advanced things, while the Prizm programmers (including myself) are writing relatively bog-standard C programs with no major tricks for speed, responsiveness, and performance, and therefore have a lot of capability still in front of them once they start to do so. I hear your defenses, though, and you should help us become a little more aware of the current state of Lua development here at Cemetech if you don't mind.

Lionel: Well-said, especially that nice word "restrictions" in there.

Of course, the Prizm will always be a step ahead because of its ability to run native code unrestricted
Also, the best Lua programmmers aren't simply TI-BASIC programmers, they knew a whole slew of other languages too


I sure will do that, and I can guarantee you that the most awesome applications haven't been released yet

For sure, but TI-BASIC is another interpreted language where you can do a lot of clever tricks to write fast and powerful programs in something that's otherwise fairly slow and limited, so I feel like the same frame of mind would carry over.

Well, that is if you want to make a smooth scrolling graphics game (Wouldn't even be possible in basic) :p

Well, I guess the best way to show the speed is making examples, and recording it through a camera (emulators are always faster).

I agree, Jim. Perhaps it would be very interesting if the best Prizm C and Nspire Lua coders got together and wrote their most optimal versions of a few sample exercises like smooth-scrolling maps and other intensive code, and see what the quantifiable speed difference is like.

As far as I see it the key issue is whether the language is fast enough or not. Lua itself is generally more than fast enough (and its widespread use in commercial game development is testament to that) but is reliant on the quality of its supporting libraries. With C you can always work around that if need be, though I hope you'd agree it would be nice if you didn't have to (though I know some people here love reinventing the wheel).

I feel a more interesting (and useful) project would be to add support for Lua to the Prizm and implement the same libraries as the Nspire's version. You could then run the same game on both devices, and if you could get it running better than TI's implementation that might give them some incentive to improve theirs.

LuaJIT makes heavy computation significantly faster, but still no match for native code - and TI's implementation is not JITed. It predates the recent (2011) release of a version of LuaJIT with ARM support
Also, I know you know that, benryves, but most of the games run on platforms significantly more powerful than our calculators are. That is, platforms where using Lua as a higher-level, interpreted language (LuaJIT is only a recent thing) helps time to market for the games, while not taking an intolerable toll on the resources.


Indeed, because TI didn't do everything wrong in their framework - but cloning TI's proprietary Lua is a huge undertaking, man-months of work...

There's quite a range of platforms and I can't find any reliable hardware specs for the Nspire but given that the platforms includes mobile phones and relatively ancient personal computers I don't think it's too unfair a comparison.


Well, maybe not all of it, but enough to form a common base suitable for game development. Their custom class library doesn't look too bad, though, with the possible excection of things like D2Editor...

Back Ground Music. Not just any, though, it's background music!

You mean like Theme A from the Tetrizm video I posted earlier in this thread? I'm afraid I haven't started experimenting with music on the Prizm enough yet to know how feasible it is.

KermM:

WOW!!! I did not see the video until right now! That's the greatest Tetris I've ever seen! Yeah, that music was perfect, especially the way it sped up! I didn't know you were so far along! Look, Mr. Mitchell: if you make it have music, I will buy a Prizm just for Tetrizm!

Haha, thanks, I'm flattered (and I'd be slightly creeped out that you know my real last name if I didn't have it plastered across the front page of the site). I don't think I'll attempt music for the first released version of Tetrizm, but when I clear out a few items on my to-do list and have a bit more free time, it definitely is worth a try.

Well, D2Editor is right now not so extensible. However, TI is going too add to the whole API some great new stuff in coming updates

@Kerm, I'm not saying that Nspire Lua is faster than the Prizm in anyway :p
My point is just that I don't want the platform to looks slower than it is

Hello everyone, this is my first post here... I own a Casio fx-CG 20, and from the first moment I have always looked for ways to expand its functionality. I have followed this forum very closely for the last weeks, and I have every single g3a available on the calculator.
I have also copied many times add-ins to my mates' calculators... free cemetech advertising!

Well, on the subject: I have seen the video demonstrating Tetrizm... in my honest opinion, it's already very functional, and I'd be pleased to give a test build a try.
You know the open source idea that says to release often, well, that's exactly what I'm asking you to do. I'm not saying to hurry making the game, I'm just saying that if you put an alpha/beta version, with what you have done at the moment, for people to download and try it themselves.

You know, other people might see bugs and come up with ideas that are better implemented now that the code is still being written and there's not much code.

But it's really up to you whether you release an alpha version or not. Just sayin'...

On another subject, I have some programming knowledge and I'd like to develop some addins myself and improve on others that have been done, the main problem is that there seems to be little information on how to compile for the Prizm, on Linux (and I don't want to reboot to Windows just to write some code!).

Cheers, and keep up the good work,
Gabriel / gbl08ma on the web

I think that Kerm might be using Tetrizm in contest #8, in which case he cannot release any WIP binaries; if he does then it will be disqualified.