Quote

If it was easy enough to code in Native (as much as in Lua with the SDK, for example), there would be a heck of a lot more people involved, and thus, the community would be way more active, including here at Omnimaga, but also at TI-Planet etc.

TI is boneheaded about limiting the functionality of their calculators (and screwing with basic user rights of doing whatever they see fit with the hardware they own), instead of taking advantage of the possibilities of native code, and their resulting attempts to "protect" themselves, is a very powerful deterrent for the open development communities. That's why even a much improved SDK would hardly help people getting involved...

Sorry but no. In our cases, since we're mostly geeky people, if we want to program in ndless, we'll go through the necessary steps to make the toolchain work etc.
But not a lot of not-so-geek people would dare to do complex stuff like this, especially if they think it can hurt their computer in whatever way (messing with the system etc.) What I'm saying here is that if it were a lot easier to program in C etc. for the Nspire, a lot more people would at least try. Example : same as the Lua SDK in 3.2, but you just write ... C code instead of Lua. How would that *not* help gaining more Nspire native programmers ?

Quote

You're too pessimistic here ;-)

Not so sure, because the fact is that hardly anybody gives a damn about the Nspire platform...

Well, if you think that way, why even bother continue talking ?
There are great opportunities out there, and TI is trying to make Lua be the easiest/most-accessible/ etc. of all.
Again, if Native coding presented the advantages of what Nspire Lua provides, it would be great. But TI doesn't want that for now. And yes, it's sad, I know, and I agree that there would be great developments made if native dev got open. But from what we've seen, you can't influence TI in this kind of decision.
So if you (particularly *you* since you're a great coder and very dedicated to the Nspire among others), decided to give up, then yes, everything will fall to pieces. "Just" because you couldn't get what you wanted from TI... That would be a bad decision because I'm sure that 3.2 and 3.3 etc. will have flaw or whatever, that the ndless team will exploit. Plus, and that's what I said before, if 3.2+ gets ndlessed, we could enjoy the best of both worlds : improved Lua API (physics etc.) and Native. Wouldn't that be awesome ?
Also : "hardly anybody gives a damn about the Nspire platform" -> We still have to think we (the whole community here at Omni+TI-Planet+Cemetech for example) do represent a very, very, very tiny part of TI's customer. There aren't many nerdy-about-calcs people out there (in high school etc. for example), as you know... So whatever we do, it will still be marginal. If, however, (for example), suddenly all the teachers went on strike to get the ability to code in native, I'd bet that TI would get their request done. Sadly for us, that's how it goes and we can't do anything about it. If the only way to make TI "move" is to release "bad" (as from TI's point of view) tools like a PTT Killer or a CAS on non-CAS (or whatever), it could only cause big troubles which could lead to the end of the Npsire "reign" (i.e : it wouldn't be accepted on exams for example, and TI wouldn't sell as much so they might stop produduction). Then we would lose everything. Would you rather have 'limited' programming (as you call it...) with expanding capabilities (Lua) and maybe even native opportunities (thanks to an eventual Ndless 3.2+) or *nothing* at all ?

Did anybody do for the Nspire what was done for quite a number of closed platforms (starting with gaming consoles), i.e. port Linux mere weeks after the release of arbitrary code execution on the platform ? No, and more than two years later, we're not remotely close to that state.
Do we have a chain of ready to use exploits, and wait for TI to close the hole of the day to release a new version, by a dozen days (often two or three days) after the release ? Not at all.
Do we have a usable graphical library (a need which was identified from the beginning) ? Several weeks ago, the answer was "no".
Do we have a program loader with relocation support (another need identified a couple years ago) ? Again, the answer used to be "no" until very recently.
This list could go on. And yes, I'm criticizing myself (as being part of a community that does a relatively bad job, compared to a number of other communities throughout the history of computing), even if I tried to do something about the math stuff, the document system, the program loader, the graphical library, etc.
Why is it that hardly anybody cares about the Nspire platform ? I don't know, but it's a fact. Another fact is that TI is (relatively) safe is a consequence of the fact we (collectively) are not trying (yet), rather than being a consequence of the fact they're making a decent job protecting their platform...

I know all that, and as a geeky-tinkering end user, I'm sad it's the case. But we can't do much about it.
Also, as you said, more and more projects of "big" envergure (english word ?) are starting/in great shape (nDSL etc.), so better late than nothing. It's not a race neither...
Also, to try to answer your question about why we're so few people tinkering with the Nspire, I think it's because ... we're working with *calculators*. Just imagine you're a total alien to this calculator-world. If you ever meet someone who tells the story of how much efforts were done to make {whatever} on a calculator, he would consider it crazy. Indeed, in a world of quad-core wifi-enabled (etc.) smartphones, who anymore gives a damn about simple "calculators". A few people. Just a few. And within that little group, a shallow (extremely tinier !) sub-group who make all the stuff and create all of what happens down there. It's a sad truth, but we really can't do anything about it. Whatever we do will be considered as a feat. nDoom on a calculator was and is, still, a great achievement. If it were done on smartphones (it probably got done), it wouldn't have gotten as much praise and source-of-astonishment than anything else. And sometimes, I guess a little-but-skilled community like ours have to face difficult times where we can't have control of everything. But you know, a lot of projects thought almost impossible at first were done on the Nspire, so let's not be depressed about that and as I said, wait&see, as 3.2+ may contain flaws that will allow us to continue with the great ndless adventure !

Quote

Just wait & see !

We'll probably see, indeed, because there always are holes. But even more than Ndless, I can't wait to see TI's dangerous strategy backfiring
It's not like they weren't warned about the potential consequences of their incompetence. The gloves have long been off, and so far, the only thing that saved them from facing the consequences of the infuriating war they've been waging against users for five years, is that nobody got over their moral bounds. But sooner rather than later, as I predicted, people are going to become interested in destructive matters (the most damaging attack venues are public knowledge)... and TI will lose, just like Sony lost, to the resent and hatred they created (and also to the false promises incompetent people at TI made to incompetent people at standardizes test regulation authorities - trying to prevent native code is in no way a requirement for certification, it's only willful and lucrative collaboration with the authorities - en français, collaborationnisme). And yet, they were warned in advance, but willfully chose not to take that into account. Their loss.
OSLauncher was perfectly harmless (and yet, it freaked out TI and the woefully incompetent standardized test regulation authorities); '2012 stuff is unlikely to be as gentle as '2010 and '2011 stuff...

Well, I talked about that up there and while I may agree on some of what you say and disagree on some others, I still don't see the point of releasing tools that could only lead to an even worse situation. Again, we are a small community trying to get great things done, so in any way, our acts won't probably have an impact on 0.1% of the worldwide Nspire Users (students who are given calcs for their studies and who only use it for their purpose : math/science), so why bother taking so much time and effort to just annoy TI ? As you say yourself, most of this community's members are not affected by PTT or stuff like this : so why fight it when we are not having to deal with it ? (This is just an example, btw, not necessarily the truth so far, idk). About the CAS-on-NONCAS thing, while this is inevitably a thought everybody had as soon as the first Nspires came out, it will again be addressed at 0.001% of the users, since most of them won't want to "take a risk" or whatever. This, we can't control...

One thing is sure, Adriweb: you'll have to take sides at some point. Either TI, or the open development community - but not both, because TI's actions are making them exclusive.

Well, so far, choosing both (or actually not chosing any side more than the other) worked pretty well : I work closely with TI with the Lua things, and I really enjoy participating in the native dev effort (but since I'm rather limited in Ndless C/(let's not even talk about ASM), although I can't do much other than testing alpha-versions of risky tools which I gladly do for the sake of going forward in native dev.).
If I wanted to take a selfish point of view but completely rational, here's what I would think :
- What profit can I make from being on the side of the open-community ? - What profit can I make from being on TI's side ?
On the first side, not much. I'm not sufficiently skilled enough like you or Extended (and many others) right now (I'll probably learn a lot more in my engineer school in the next 3 years) to create incredible things like you do, and leading to a great fame (on the Internet). And since TI knows me and has (with others like Levak and Critor for example) taken some never-seen-before close-contacts with me/us, it would destroy all this (since they wouldn't probably (?) place any trust in me anymore). So, from that point of view, I just lose everything I'd have gained after so much time of hard work.
If however, I place myself on TI's side, I don't see any bad points, actually. I don't think (as long as I'm not an TI employee) there are rules that forbid being part of a perfectly legal community like Omni/TIPlanet. So then I can profit from both sides : while having the advantages of working with the latest TI stuff and having their trust, I can also continue what I've always done since I bought my first TI : making programs and messing around with what I can. More recently on the Nspire, I focused myself more on Lua since it's where I'm good at, and I didn't have time to properly learn C (and ASM not at all) to be able to code for the Nspire with Ndless. As said, I will probably have more time next years to learn specifically that... You can take Levak's case (which is pretty close to mine, actually) for example : since he's better than me at C/ASM, he does great stuff too for Ndless. But he's great too in Lua so he also has the opportunity to work with TI, as close as me. And that was thanks to "official" programming (well, we forced it because we were tinkerers who allowed Lua programming before it was released). Now we have job opportunities etc. Sadly, this isn't what "underground" Native development could have done for us. Although TI would gladly hire you and you know it ;-) (but this is your choices so I don't have to say anything about what you want to do ) If I'm given the opportunity to work for TI when I get my engineer degree, it will be a great achievement for me, since it would mean that after years of messing with calculators, I'm finally "recognized" by the company, even if it wasn't my goal at all at first (I was just messing around for fun, but then I got better at programming etc.).

But I don't [ want to ] think like this. What I think at a long term would be better for me (and probably a lot of people), is to try to focus on what TI can offer, since it's pretty much the only way not to be constantly fighting stuff that would have to be redone for each version. As you said yourself, TI will probably block Ndless in 3.2, and a lot of work would have to be done by people just like us who give their free time for a cause like this. I'd be more than happy to participate more into native dev if TI made it easier, like at least not blocking everything each version. And a lot of people are be probably like me, on this point at least.
So, so far, I'm not having to chose a specific side, and I wish it stays like that as long as possible, so I can participate in the great "official" stuff, while also enjoy the incredible feats of a devoted community (as a end-user and also a participant as much as I can find myself helpful to help the programmers with testing etc.).


I would have written more and mroe to explain my ideas, but it's getting late, and I don't want to post stuff that take an hour to read (oops, that might be too late )

Quote from: DJ_O on April 06, 2012, 08:42:57 pm

Yeah. Also there's Lua, but many people find it too slow to perform any advanced stuff. As for OS updates, sadly the issue is that retail store calcs will eventually all come pre-installed with OS 3.2 and a lot of teachers force students to upgrade.

I think you're mistaken. Or at least don't understand everything going on. (nothing bad intended towards you)
Of course C/Asm is faster. We know that, we get it.
But you can do great things with Lua. Just look at awesome stuff already out even before Physics is out and before the official SDK is out.
Can you imagine how even better Lua scripts are going to be when all that is released ?
Compared to what the Nspire always had (crappy Basic), Lua scripting is infinitely better.
Another thing to note : Everybody knows that Basic is slower than native. Look at popular z80 and 68k calcs. Count how many BASIC programs there are out there. Do you think they are all crappy because it is "too slow to perform any advanced stuff". No.
So, what we have right here: a "limited" basic, but making incredible things thanks to everybody in the community.
Now, Lua on the Nspire allows us to make 9001* better things that what z80 and 68k Basic could allow us to do.
So please, stop writing again and again that Lua is so slow we can't do a thing with it. (this is not necessarily to you but to everybody in general, I mean).
And also, while you may want to compare raw computation speed, you will compare things with criteria that won't be fair for both the compared languages. For example :
  You can't make a NES Emulator in Lua. You can in native.  Native Wins.
  You can't make a GBC Emulator in Lua. You can in native.  Native Wins.
  You can't process xxxx computations per second in Lua.   Native Wins.
  You can't access USB over Lua.  Native Wins.
Clearly, Native wins right ? Lua is crappy right ?

Now let's take this point of view :
  Nspire Lua is easy to code in. Start up TINCS and open the Lua SDK. It's a mess with Native.  Lua wins.
  Lua can access the math engine. (with CAS if device is CAS).  Native can't.   Lua wins.
  Lua can interact with the already existing variables and widgets.  Native can't.  Lua wins.
  Lua is supported by TI and whenever the SDK is out, so it will have an even bigger community.  Native isn't/won't (as much).  Lua wins.
Clearly, Lua wins right ? Native is crappy right ?


As you can see, you (or any other) can't just claim that "it's too slow". For many things, Lua's better than Native. And for a lot of things I'm pretty sure it will stay that way. Except if TI radically changes its ideas and publishes some Native Resources Toolkit or whatever, but I don't see that coming soon...

Also, some other things :
- I think the community needs to make some kind of all-in-one-included Native SDK, where you could just select the target device, and everything would be so you just have to code your thing, and it would compile with the latest ndless toolkit available (or the one you want if you want to override that), etc. so that people who aren't really geeky enough to setup properly their toolchain onto their system (or just dont want to mess with it too much, or for any other reason), could finally be able to program in C for their device ! If it was easy enough to code in Native (as much as in Lua with the SDK, for example), there would be a heck of a lot more people involved, and thus, the community would be way more active, including here at Omnimaga, but also at TI-Planet etc.

- While you can think 3.2 is blocking ndless (it will most probably do), there will probably be flaws like in every other OSes, so don't worry too much I guess

Yeah but I mean for example on the 83+ you got a slow but easy to code language, along with a fast but harder to startup with (and hard to code when it comes for certain type of games and stuff like math softwares). On the Nspire, you got something in the middle, probably more on the easy/slower side by default. Basically you got less freedom by default and there is always a possibility (the worst case scenario) that ASM/C might never be possible again on the Nspire. If that ever happened (I doubt it will and if it does, eventually a later OS might re-open a hole allowing Ndless to run again), then you still got a nice language called Lua, but the issue is that many members visit us for programs like emulators and Doom. Let's say for example that the best mario clone that could ever be made in Lua has 10 keypresses per second maximum and a framerate of 20 FPS (which looks like 10 cuz of the slow getkey). I'm pretty sure this won't be a huge "seller" like SMB3 on NESpire, simply because it isn't as good gameplay-wise as a real Mario game or even Sam Heald's. Basically, what I am saying is that we will attract much less new members and visitors with Lua than we did with Ndless. Thankfully TI has made Lua easy to get started with, so that already helps us a lot. I just mean without Ndless this could still hurt us despite the presence of Lua, and without Ndless, eventually people will have pushed Lua to the limits and run out of ideas for new innovative Nspire games.

Basically my post above was in function of the worst case scenario about Ndless. Do you think TI-83+ calcs would ever have gotten this popular if ASM was not possible on them? Check the download stats: Super Mario 1.2 is among the top downloads since the past decade.

Quote from: DJ_O on April 06, 2012, 08:42:57 pm

In the worst case scenario, I expect 90% of the Nspire scene to be european developers in a few years (since in Europe teachers do not have full control of what students do with their calc).

Maybe I haven't understood correctly, but how is that bad ?
I mean, 90% of the "underground" scene is European ?
on the Nspire, 3.2 does also mean Lua physics stuff, so even more content...
And that could be from anywhere in the world

You effectively misunderstood, by far. I stated this because in United States of America, most schools force students to upgrade their TI-Nspire calculator to the latest OS in order to be allowed to use them. This is not the case in Europe. I was saying 90% in reference to how almost everyone in USA would no longer have access to Ndless in the worst case scenario.

- TI isn't that evil either. I recently asked them why they removed/disabled the print command in Lua. They reason was not to annoy us, and now there is a very big chance that it will return in the OS after 3.2.

Basically it might just have been a mistake?

Hopefully they don't screw up too much either like they did with OS 2.53MP breaking many ASM programs and new 84+ hardware preventing RealSound and some other programs from running. X.x

I think changing the theme to a lighter, more modern and less "intimidating" one would encourage others to participate.

Hmm, maybe changing the theme so that the preference is saved? (or maybe it is and my cookies are getting cleared or something)

However, I heavily suggest developing simultaneously on the PRIZM too, to give TI more competition by trying to attract PRIZM users.

What needs to be understood, is that we (the TI-Nspire community mainly) are minor minority. TI does not give a damn about us, that's the harsh reality of things but it's at the same time something very normal and comprehensible. TI's a massive corporation, and we're just surface dust next to shareholders. As I said earlier, that's how business works, that's how you keep the cash flowing. There might be some patting going on and fairy tale ideals floating in the air, but we have to be realistic: unless we are the ones to strike, we will have no effect, nada. We're just not enough for passive actions to have any kind of noticeable effect. Even if every single person on Omnimaga switched to the Prizm, it would only be a lonely, insignificant fart in the Milky Way.

Quote from: DJ_O on April 07, 2012, 06:59:06 pm

However, I heavily suggest developing simultaneously on the PRIZM too, to give TI more competition by trying to attract PRIZM users.

What needs to be understood, is that we (the TI-Nspire community mainly) are minor minority. TI does not give a damn about us, that's the harsh reality of things but it's at the same time something very normal and comprehensible. TI's a massive corporation, and we're just surface dust next to shareholders. As I said earlier, that's how business works, that's how you keep the cash flowing. There might be some patting going on and fairy tale ideals floating in the air, but we have to be realistic: unless we are the ones to strike, we will have no effect, nada. We're just not enough for passive actions to have any kind of noticeable effect. Even if every single person on Omnimaga switched to the Prizm, it would only be a lonely, insignificant fart in the Milky Way.

Although on the other hand, developing simultaneously for both the PRIZM and Nspire like Reo is doing gives you a larger audience for your programs (or reduces the effects of a permanent Ndless lockup on your audience)

Yeah I guess that part could be moved out, although I guess it's inevitable since the biggest part of Omnimaga activity that got affected is Nspire-related.

Casio has expressed concerns about 3rd party developement before and have no plan for a SDK. They didn't expect add-ins to be made by the community, but they have decided to give us one chance and they won't block them unless we start publishing softwares such as image converters for the american models, tools to cheat in tests and stuff that can permanently damage the calc.

Also Alberthro I know Lua isn't like BASIC, I was saying it's kinda like between ASM/C and BASIC. Not too slow and limited, but not offering 100% speed/freedom either. Basically, good enough to make somewhat cool games, but not enough to pull a gbc4nspire in Lua that will draw hundreds of members to Omnimaga, TI-Planet and ticalc.org whenever Ndless is useable on the latest OS. More programming examples and routine examples (such as map engines, collision detection and data management) could possibly improve Lua audience, though, because the language, while being limited, is kinda underused lately.

Anyway if we split this discussion I guess it could be called The Future of TI-Nspire programming. The debate above somewhat remained fine with no trollish message towards an entire brand of calculator except maybe one, so it could definitively continue on in a different thread and we could possibly use some more of those calc-related debates (at least they're more constructive and interesting than political and religious debates).

EDIT: Wow I was scared Omni went down again as I clicked Save and nothing happened. Glad it was due to a topic split.