Flash Archive Undelete... - Page 1

Sur le forum de TIGCC/TICT ( http://tichessteamhq.yuku.com/topic/4626 ), Travis Evans a posté un programme qui s'occupe de récupérer des variables archivées effacées, sur 89/89T/92+/V200 tournant AMS 2.00 et plus.
Il a uploadé source (GPL) et binaire compatible on-calc à http://members.cox.net/travisgevans/ti-calc/undelete.html

Naturellement, il lui faudrait plus de beta-testeurs, et pas tous ceux qui fréquentent ce forum n'ont un compte sur le forum de TIGCC/TICT (et parmi ceux qui en ont un, je doute que tous consultent tous les quelques jours le forum de TIGCC/TICT ). Je fais donc un topic ici.

Travis Evans ne parle manifestement pas le français. Par conséquent, afin qu'il puisse utiliser directement la plupart des reports et suggestions qui lui seront faits dans ce topic, exprimez-vous SVP en anglais (à moins que votre anglais soit "terrible", auquel cas il vaut beaucoup mieux s'exprimer en français, et laisser d'autres _humains_ traduire ).

tiens, tiens

Oui, je sais
A ma connaissance, Flash Savior n'est plus maintenu et/ou ne fonctionne pas sur 89T, n'est-ce pas ?

Il y a une version 2 qui n'est jamais sortie (je crois.)
Elle est limitée à AMS 2.05 de toutes façons (à cause des crédits dans la barre de calcul surtout (), et aussi peut être de l'adresse du premier secteur de flash)
Je pense que ce sont les seules choses à adapter pour les calculettes récentes.
Encore que je me demande si les crédits ne passent pas dans le mode *alternatif* avec un AMS inconnu... M'enfin il reste de toutes façons la MAJ de l'adresse du premier secteur à faire.

It doesn't work with PedroM 0.82
I get "AMS 2.00 or higher needed" in the status bar, and I am pretty sure I have the most recent version of PedroM.

PedroM 0.82 isn't AMS, so it's definitely not "AMS 2.00 or higher".
Isn't PedroM's archive format different from the AMS one in the first place? Or does the program actually work if you hack around the AMS check (using NO_AMS_CHECK)?

Without AMS version check, PedroM gives "ROM call unavailable" (it doesn't say which one, though). But if the needed ROM calls were implemented, PedroM would have to use exactly the same archive format as AMS 2 and higher, because Undelete totally depends on this.

Strangely, after attempting to run the program, I also can't receive any more files (in TiEmu, at least) until I reset PedroM.

Congratulations Travis

Thibaut (./8) :
Congratulations Travis

PedroM gives "ROM call unavailable" (it doesn't say which one, though)

Probably ROM_CALL_3CF, which I dubbed "EM_ArchiveMemoryBeginning". I left this name instead of the official one given by Scott Noveck in http://tichessteamhq.yuku.com/topic/2385 , since the official name of this ROM_CALL comes from TI's private header file.

If ROM_CALL_3CF being unimplemented is the only thing that prevents your program from running on PedroM (i.e. that the archive format of PedroM is 100% compatible with that of AMS 2.xx), that ROM_CALL can be easily implemented in PedroM. Assuming the archive memory starts at the nearest sector boundary following the end of the OS, the value it returns should be something like
(((ROM_BASE + 0x12000 + __ld_program_size + <several dozens of bytes for the certificate fields>) + 0x10000) & ~0xFFFF)

The archive memory format is different anyway, so the program won't work properly on PedroM anyway. PedroM is just not supported, we'd need a special version of the program for PedroM. As it's GPLed, as is PedroM, anybody interested can implement that, so I don't see why you're nagging Travis for it.

Strangely, after attempting to run the program, I also can't receive any more files (in TiEmu, at least) until I reset PedroM.

The PedRom's author reported to us this problem. If I remember well, it's a PedRom related problem, not a TiEmu one.

Okay, thanks, everybody. I made a small page at
http://members.cox.net/travisgevans/ti-calc/undelete.html and posted a preliminary copy of Undelete v0.01.00 on it. It's preliminary because I haven't yet tested it as much as usual, so be careful and back up your calc first. It's a small update consisting of mainly UI improvements that probably aren't very noticeable unless you know what to look for. The included change log lists all the details.

This new page should now be used instead of the temporary link that Lionel posted earlier.

I have edited post #0.