KhiCAS, le moteur formel Giac/Xcas pour toutes TI-Nspire

[Adriweb]

Interesting...
Regarding the status bar, it's something that could probably be "fixed" (force full refresh, rather) on the Lua side (platform.window:setFocus(false) platform.window:setFocus(true)), or the C side, it shouldn't be an issue.

I see a new boot loading screen that I have never seen before (or maybe my memory is failing me... but mine shows this, for instance)


Could you tell us what hardware revision it is ?
This is the last character of the serial number at the back of the calculator (actually, if you could give the whole group (like, X-XXXXX), it'd be nice.
Thanks.

[Strontium]

P-1214T is my revision.

The different screens could be because yours is a CAS and mine isn't, but I don't know for sure.

[Adriweb]

Nah, in this case, a recent CX CAS would should like yours + CAS, I suppose.

But that still doesn't really explain the weirdness seen in KhiCAS, however... Well, new HW-revs could have specific behaviours that Ndless doesn't get on well with, but... we certainly don't know much about it so far if that's the case :/

[critor]

Hi Strontium,


Indeed, something's strange or new with your calculator.

Could you please dump it for us by sending and running the Ndless program Polydumper :
archives_voir.php?id=3829

It should generate some images in the ndless folder :
- manuf.img
- boot2.img
- diags.img
- os.img

Could you get and share the first three ones with us ?


And what's the "P-0000X" datestamp printed on the back at the right of the serial number ?


Regards.

[Bisam]

The datestamp on the back says "P-1214T"

Heres the output of PolyDumper:

polydumper.zip

(1.11 Mio) Téléchargé 73 fois

Also, could you explain what boot1.img, boot2.img, diags.img, and manuf.img are?

[Lionel Debroux]

* boot1.img is the first stage of the bootloader, which takes care of the initial display, advancing the progress bar to half, handling the maintenance menu key combo, etc.
* boot2.img is the second stage of the bootloader, which takes care of loading (decrypting and uncompressing) and launching the OS;
* diags.img is the diagnostics software, it contains code for interactive testing of the keyboard, screen, etc.
* manuf.img is the manufacturer data, which contains various pieces of information about the calculator, including the data structures used for displaying graphical content in the boot1 and boot2.

EDIT: now for some analysis of your files:
* your boot1 bears the 3.0.99 version number, nothing new;
* your boot2 bears the 3.9.1.34 version number, nothing to see here either;
* your diags bears the 3.9.1.37 version number, ditto;
* I'll let others analyze the manufacturer data.

[Vogtinator]

I've seen that screen somewhere, but I don't remember where.

The manuf.img has bit 7 set at 0x81C, probably indicating a CXCR.
It may be that the compressed background image is a different one, I don't have a flash image here right now to compare them.

[critor]

Thank you very much Strontium.


Yeah the manuf had a new background, but the other files are normal.

I don't use OS 3.9, only 3.6 - which may be why I don't have such problem.

[parisse]

Je viens de mettre en ligne la nouvelle version de khicas pour les nspire CX:
http://www-fourier.ujf-grenoble.fr/~parisse/calc/khicas.zip
Avec donc des flottants multi-precision (par ex. evalf(pi,30)) et de l'arithmetique d'intervalles (par ex. p:=[1.56..1.57]; exp(p)).
Il reste surement pas mal de bugs, mais j'espere avoir supprime les plus evidents...

[Strontium]

That update fixed both of the issues I have been having!

Edit: A new bug has appeared. Clearing the history [menu>B>8] causes a crash. Not a huge deal since I don't use it often, but it exists. Also, the original crashing issue appears to still exist. When I went to see if it was still there after updating, I was able to open and close KhiCAS twice without having it crash. No clue what was going on there.