Posté le 24/10/2020 11:36
Planète Casio v4.3 © créé par Neuronix et Muelsaco 2004 - 2024 | Il y a 99 connectés | Nous contacter | Qui sommes-nous ? | Licences et remerciements
Planète Casio est un site communautaire non affilié à Casio. Toute reproduction de Planète Casio, même partielle, est interdite.
Les programmes et autres publications présentes sur Planète Casio restent la propriété de leurs auteurs et peuvent être soumis à des licences ou copyrights.
CASIO est une marque déposée par CASIO Computer Co., Ltd
Citer : Posté le 24/10/2020 11:38 | #
Here is some documentation on the g1m format
Citer : Posté le 24/10/2020 11:42 | #
Here is some documentation on the g1m format
Excellent, thanks! I see there are multiple types possible, basic programs, listfiles, pictures, captures. I already know about basic programs, but is there more documentation about the other files? This certainly helps me a lot!!
Citer : Posté le 24/10/2020 11:45 | #
I can't find anything else on the PC bible, though someone more knowledgeable may be able to give more documentation.
Citer : Posté le 24/10/2020 14:27 | #
Hi Pt_.
In exchange, would you contribute to the documentation of Casio file formats ?
.g1m images, .g3m images, .g3p... everything supported in SC.
Casio file formats aren't very well documented, and it's a pity I can't find anything public about the reverse analysis done for the support in SC.
Citer : Posté le 24/10/2020 14:29 | #
Hi Pt_.
In exchange, would you contribute to the documentation of Casio file formats ?
.g1m images, .g3m images, .g3p... everything supported in SC.
Casio file formats aren't very well documented, and it's a pity I can't find anything public about the reverse analysis done for the support in SC.
Well, sure, I'm already very hard busy reverse-engineering the formats, and it certainly will be used in SourceCoder, plus documented elsewhere
Citer : Posté le 24/10/2020 14:32 | #
Thanks.
Can't you start by sharing what has already been discovered ?
Since SC already supports .g1m, .g3m and .g3p files.
Citer : Posté le 24/10/2020 14:35 | #
Sure, here you go: https://pastebin.com/gMLmbiH1. These are the formats for 3 types: spreadsheets, programs and equations. I will just continue working on this
Citer : Posté le 24/10/2020 14:37 | #
Thanks.
How about .g3p files, and .g1m/g3m pictures ?
They're supported by the SC converter.
Citer : Posté le 24/10/2020 14:39 | #
They are supported yeah, but that doesn't mean I know the format :P when I know more I will post it too!
Citer : Posté le 24/10/2020 14:48 | #
Oh @Cakeisalie5 put a great deal of work into reverse-engineering these formats. He's not actively developing anymore but there is still a lot of info out there. You might find significant details in the libcasio repostiory, especially in include/libcasio/format.
I know he discussed some details of the formats with Simon Lothar on Casiopeia so if you need something obscure and yet undocumented, you might have luck searching the forum there.
Citer : Posté le 24/10/2020 14:49 | #
@Pt_ Thanks. If they're supported, then someone must know about them.
When I released my image converter for the fx-CP400/CG500 years ago, I publicly released my (probably very incomplete) reverse engineering of the format.
Leaving public traces of the file format reverse engineering is very important as it'll benefit future developers of conversion tools, linking softwares or emulators, and thus the whole community.
Citer : Posté le 24/10/2020 16:29 | #
Started with documenting: https://docs.google.com/document/d/18nVDz1mtYw04odErdSzvp_x5GSkAYppmcU6cf3NxU9k/edit?usp=sharing
Have fun!
Citer : Posté le 24/10/2020 16:31 | #
Thank you very much.
Citer : Posté le 13/12/2020 16:48 | #
Just saw the efforts provided here. Very nice!
Just to explain kind of the logic of the g[1-3][mp…] files:
- a standard headers, 32 (0x20) bytes long, obfuscated (structure once un-obfuscated, with correspondances to identify the file type beyond the standard header).
one exception: the storage format, g1s, which contains just a lot of blanks until the address in SH memory (0x02700000 or something i don't remember) then the raw storage memory.
Then if it's a g3* thingy, it has a standard subheader, described here with some correspondances defined here
Then specific subheaders, such as the G3P subheader, then the zlib-compressed R5G6B5 picture data and an adler32 checksum and some footers i think, i don't remember exactly for those
G1M and CAS files are main memory archives, with files but not with the usual metadata. G1M are the "modern" format (as opposed to CAS40/CAS50/CASDYN which was very linked to how they were represented in the legacy protocols). Main memory is called "MCS" by Simon Lothar, for Main Control Structure, and some formats are documented here, with the general G1M/G1R/G2M/G3M/G3R format given here (there are some subtitilies, e.g. file "groups", when parsing), the correspondances between what is in the header and the decode function to use is here (quite a few as you can see, and it's not complete at all).
There are many more things in these files, and I was also very interested in legacy formats, but ended up giving up in 2018 because i'm not organized enough. Feel free to hit me up with questions if needed
Promotion ordinaire sur les inscriptions sur
Mon blog ⋅ Besoin d'utilitaires de transfert vers et depuis la calculatrice sous GNU/Linux ?
Citer : Posté le 13/12/2020 17:20 | #
Pendant qu'on en est là, est-ce que tu peux résumer l'état de mcsfile ? Quelles parties sont utilisables ? Quels formats sont décemment complets ? Si quelqu'un voulait reprendre le projet, où est le code à étendre ? Haut niveau (description des formats) ou faut descendre compléter les APIs internes/etc ?
Citer : Posté le 13/12/2020 18:14 | #
Très bonne question, à laquelle je ne peux pas vraiment te répondre parce que ça fait longtemps que je n'ai pas continué ça. En fait ce qu'il s'est passé c'est que j'avais fait des éléments d'I/O très généraux dans la libcasio, qui avaient commencé par une simple utilisation de fopencookie() dans p7, puis que j'ai isolé dans la libtio et c'est en voulant développer la libtio que j'ai stall. La première chose que Lailoulezz a fait en clônant la libcasio c'est remettre ces éléments dans la libcasio pour simplifier, et il a probablement raison, ça peut être une idée de repartir de sa branche.
Après un jour je referai peut-être ça dans un langage plus moderne, où je n'aurais pas envie de refaire un équivalent de la libtio qui m'a stall le projet. Je me tâte à essayer Rust avec nom, ça peut être un projet plus riche mais je m'écarterais alors du C et de sa compatibilité. On sait ce qu'il en est du support du SuperH dans LLVM ?
Promotion ordinaire sur les inscriptions sur
Mon blog ⋅ Besoin d'utilitaires de transfert vers et depuis la calculatrice sous GNU/Linux ?
Citer : Posté le 13/12/2020 18:25 | #
À mon sens, la valeur du projet est supérieure à ses choix d'implémentation - si quelque chose peut te motiver à y retourner, lâche-toi.
LLVM n'a pas (et semble loin d'avoir) un backend SuperH : http://llvm.org/docs/CodeGenerator.html#target-feature-matrix
Citer : Posté le 14/03/2021 22:34 | #
@Pt_
Did you notice this post about the g3p and g3b file formats ?
https://www.omnimaga.org/casio-calculator-programming-news-and-support/hacking-into-the-g3pb-format-(halted)/