I have noticed that prizm development overall appears to be infrequent and comprised of a few developers and is not something that is done as much as, say TI calculator development at-least the ones were TI somewhat lets you control the device you purchased. I understand that they have a larger market share (not due to technical superiority) but I am still surprised in the lack of prizm development as, (I find anyway) that it is an easy platform to develop for. So I am wondering what do prizm developers want? Already we have good documentation on the prizm wiki, example code, libfxcg which is getting better almost on a daily basis and GCC support the sh4 target so it is easy to compile programs. I am wondering if there are any prizm users here who would have liked to make a prizm program but had issues doing such? These issues may have since been resolved or be something that I can resolve. I can say right now that after I do some smaller projects for the prizm I plan to work on a larger project. I think it will be a port of TuxMath but I am not sure if that is the right direction to take. I want to make something that will make a large positive impact on the community and I am not sure what to do. I am wondering what will really bring interest in the platform. I wonder what do the prizm users want to see?

I agree with you about the Prizm-developper community. But, Prizm developpement isn't completely dead, what I see in Planète-Casio : we are actually in the 10th Birthday Contest, and this contest is open to Prizm entries, actually 3 Prizms addins (but they're still quite secret, but here's a pict of one) are planned and 1 Basic game (a courageous programmer ^^), the contests 2nd prizes are actually 3 Prizm (1 for solo participation, 2 for the duos), and I imagine that the persons who are going to get them will program them quite well (as they'll be 2nd of the contest) .

We also have a guy who's developping a SDKolor (http://www.planet-casio.com/Fr/forums/topic13082-1-SDKolor.html) which is here to encourage beginners to program the Prizm (even if I'm not completely right with the software itself (not open source, quite big or not compatible with all the versions of Windows, but, as it's only for Windows, I'm not concerned about :p ).
There's also an sdk for Linux (actually for the Monochrome's Casio) which could easily be ported to the Prizm Platform, the author is actually thinking about that.

I think that's some good initiative that could profite to the community .

The first thing that I would recommend (and is already going on) is a good, stable and clean libfxcg.
About the documentation, there is a very good amount of work already done, that's so nice !
If anything related to the setting up of a dev environnement is up to date, I don't see anything to suggest right now, but I'll talk a bit with the French community of this topic .

Most of the English-speaking Prizm community is here at Cemetech, so unless you wanted it to be a French-only contest, I think advertising this (and future) contests here at Cemetech would probably be a good idea, and of course encouraging users to submit their completed projects to our archives. The more visible software there is for the Prizm, hopefully the more users and programmers the platform will get.

Yes I agree with KermMartian. I did not even know there is a contest and of course wish I did. Can you give more details? Can I even enter?

please continue using prizm , you believe that few people use it but the few peolple who use it are you grateful to your effort

Ill be joining the ranks of Prizm coders very soon (my prizm is expected to (finally!) arrive this Thursday). While I cant yet develop anything, I have tried looking into what it takes to develop, and at existing programs available. however, one thing struck me as scary: the multitude of prizm bricking I seemed to come across (so it seemed to me, anyway). perhaps something to help absolute beginners out, put my fears to rest, and serve as a leaping board to broader prizm development would be fairly nice.

I think the bricking was almost exclusively due to overclocking, wasn't it? I think if avoid that (or addins that do that) you should be okay.

I think there were some special areas of memory that caused trouble too - but gbl08ma seems to be the expert on that.

Anything that erases, or causes the erasure of, the first flash sector, will brick the Prizm. There may be other causes of bricking, too - nobody ever connected a flash reader to the bricked calculators, to find out what is the state of the flash. The only case where we're sure bricking was due to the bootloader disappearing, is with my old calculator, where the first flash sector was erased but nothing else was damaged and the calculator kept working until it was rebooted. See http://prizm.cemetech.net/index.php/CASIOABS#Behavior_on_Damage for more information.

It's hard to find concrete proof that overclocking is involved, but it may contribute to flash damage. I have seen that for those calculators which can't overclock to the higher speeds, the problems seem to be related to flash read/write.

Well with that said the theoretical possibility of an un-bricker may be be a possibility making this issue a thing of the past. What I mean is something that can fix the bootloader as it is evident that the bootloader can be overwritten so such a task may be recreated to write predictable data which is the default bootloader. I am afraid that I am not brave enough to test such theory. I do not want to brick my calculator while trying to make a program to do the opposite. Anyway if I ever get a good method to desolder the flash chip without breaking my calculator I could socket the flash chip and get a nice programmer and do some dangerous stuff and just restore the flash chip if I mess anything up. Anyway Luxen it is good to hear that you will be a prizm programmer soon.

0o, sorry, I thought I had already aswered here :s...

So, talking about the contest, he's, I'm sure, open to non-French entries, but from what I see in the rules, the prizes can only be send within the European Union (more some countries like Switzerland...). I'm not sure exception would exist, if you're really interested in winning a physical prize, perheaps you could see with Totoyo, the admin who pay for the shipping costs, but I'm not sure of anything.

By the way, you're for sure allowed to participate without considering winning a physical prize. You can participate in solo or in duo, and can program an addin or a basic game for the Prizm or the monochrome Casio Graph (I don't remember their "internationnal name", fx-98*something*). The contest is actually ending on november, 2nd at 23h59 (French time). You'll have to upload you're game on Planète Casio, but if there's some people interested in participating, I can try to explain more "formalities" .

Oh, i almost forget that the theme of the contest is : "Adventure" (pretty large one :p )

I am quite surprised that our admins didn't advertise here by any mean...

And talking of overclocking, if you program carefully, you can do some kind of games that runs smoothly without any "risky stuff" : for example, my entry is actually running at 23-26 fps (I'll lost perheaps some in the future because of some things more complex being added, but the main engine is running) without OCing for a "pokemon like view" (but without scrolling) : the forgotten syscall Bdisp_PutDisp_DD_stripe and some sorting stuff can do a very good job ! I'm sure that GravityDuck (for example, because it's some kind of reference in Prizm game dev and use a built in OC) could win at least 4-5 fps using a similar structure, so that would make OC a bit more useless

I am not sure how much shipping cost are but don't you have to pay a huge tax and I am not sure how much shipping would be in addition. If it were much higher than to another European country I would feel a bit bad. Also I did end up looking at the contest using google translate and the post said the fx-cg10 is supported which implies that American entries are welcome but yes I would not be surprised that shipping would be an issue and it would feel like leeching of the community as lets say I do win a calculator I would of course like to contribute to the community but I cannot speak another language besides English.

So you are saying that I just have to pay attention to programs that overclock the calc? supposing it is mentioned in the brief of the program, this seems simple enough.

Have you considered detecting the current clock speed and if it is not overclocked already display a message on screen asking if you want to overclock while the programming is running and if the pressed yes remember to restore the settings when done. If the calculator was already overclocked leave it is as when exiting the program.

I am trying to learn how to program for Casio FX calculators but found it hard to locate all the resources or simply don't know where to place them. So some better and more complete tutorial with live not dead links would be great help. I tried compiling some examples available here using make.bat method with makefiles but it seams really hard to get everything in place properly: for example which mkg3a.exe supports eActivity strips and what's the best place to put it relative to the example project folder please? Also my setup refuses to compile with -flto under LDFLAGS in makefile, same with std=c++11 and D__GIT_TIMESTAMP="$(GIT_TIMESTAMP)" under CFLAGS or -lceigen -lfromcasio under LIBS. I have not figured out unfortunately where to find them and where to place which for the compilers to work so resorting to learning from more basic examples but would like to give more complex ones a go too, if someone could kindly direct to where I can find the required downloads and instructions. Many thanks in advance for your help.

That is specific to gbl08ma's programs. If you do not know what those do you do not need them. If you want to know I could write something but I do not want to further confuse you. Try this simple makefile
https://gist.github.com/ComputerNerd/4df44359dcf85d1a2320
Be sure to replace names with what you are doing.

Thank you very much for your help - I will experiment with this.

I have been trying to compile newest libfxcg from https://github.com/Jonimoose/libfxcg and https://github.com/gbl08ma/libfxcg for Windows and failed every time. I am using makefiles which came with above. Does anybody know how to fix this on Windows please.

Many thanks in advance

Well you should be able to quickly notice that it is trying to run a program that you do not have called casio-gcc. You need to fix that. To do this in the folder called libc edit the makefile

Code:

With

Code:

Now do the same for the makefile in the folder libfxcg replace

Code:

With

Code:

Thank you very much. I ended up with a couple of libc and libfxcg folders - a set under PrizmSDK-0.3 and another under projects folder all with makefiles similar to yours (pmed already) now but still had problems compiling. Will go through it properly tomorrow and let you know of any problems. Many thanks for all your help.