Jump to content
Jagware

SebRmv

Level1
  • Content count

    1,553
  • Joined

  • Last visited

  • Days Won

    1

Posts posted by SebRmv


  1. No it's rather small. the biggest was like 70k, mostly like 40k though.

    However this latest binary doesn't seem to write a file anymore...

     

    There is one thing that came to my mind.. in the skunk lib it says, no of the commands are "thread safe" (see skunk.s)... Means, maybe the code is messing something up regarding the encryption tool or the data? Just an idea.

     

    Now that I think about it, this new binary you posted (the one for the skunk), behaves strange compared to previous ones.

    Sometimes it closes the console, sometimes it used to create a file (but maybe that was just luck?) and now when I try it, it ends with a bright green screen on the jag side after the calculations and no response on the pc side, waiting for the 68k to handshake with the console, as if the jag crashed or so

     

    What did you modify to previous versions?

     

    Almost nothing was modified. I just moved a bit the code to gather all the skunkboard related code.

    I also added two skunkNOP.

     

    Well, that's very strange. I will contact Matthias to ask him whether I interpreted correcly his code.

     


  2. I spent lots of time travelling by train and playing katomic, and I actually loved it.

     

    That's how I got the idea to port it onto the Jaguar in 2006 as I thought it would be a very good "first try" program.

     

    The Reloaded version shares absolutely no code with the previous version. It is entirely written in C using my library.

    (well except for the plasma... and the hidden/final screen ;))


  3. I just did a few tests myself using the transfer code and it should be fine.

    My only guess is, something is messing up the data before tranfser.

    Maybe it's the uart code causing trouble, maybe it's something skunk related causing trouble with the jag cd (although I doubt this, as the rest of the program seems to run fine)...

     

    can anybody test if the data using the xmodem cable is correct or corrupted as well?

     

     

    There should not be any interactions with the UART code, since it is completely deactivated.

     

    I tried apart the transfer code and it works well. So this can only be a bad interaction between either the JagCD and the skunk or between the encryption code and the skunk library.

     

     


  4. Unfortunately, I don't have at home the required hardware to test my skunkboard version of the encryption process.

    If anyone is volunteer to test it, I'll be happy to send the cof file.

    Otherwise, I'll wait to get hands on a stock JagCD unit (I only have my dev JagCD at home, and it does not seem to work with a stock Jaguar)

     

     

     


  5. Thank you very much :-)

     

    Also thanks a lot for offering your help. Personally I didn't get very far yet, so your help is very much appreciated. :-)

    Do you know if the data sent back to pc can be written to a file? or just the console window?

    If you want to discuss this privately, I'm happy to talk via PM or e-mail :-)

    I think it would really help the community to have a usb based encryption approach.

     

    Regards, Lars.

     

    No problem to discuss about this on this forum.

     

    My modified version of JCP emulates all system calls related to files, so that I have a standard interface on the Jaguar side to access files on my PC (running linux).

    This thus includes writing files.

     

    I had a quick look to the encryption program (modified by Mathias Domin to run on BJL) and I think it should not be too difficult.

    The main "difficulty" is to split big transfers into small ones, because Skunkboard may transfer blocks of data of about at most 4ko at once.

    (so for bigger blocks, just split them into 4k blocks and take care)

     

    I think it would probably be a good idea that the encryption process works on the original JCP version of Tursi (ie not on my modified version).

    If I remember correctly, it should be possible to write files with it. (I will double check that)

    However, I fear that there is a limitation over the size of the files (typically limited to one 4ko block).

    Is it sufficient for the encryption key? What is its standard size?

     

     

     

×