Hello,
I got the egate driver  from Muscle web site and used it  with
e-gate Cyberflex Access 32K card:   ATR= 3B 75 94 00 00 62 02 02 02 01

This card  is codeshield enabled.  As you say, .cap file must be post processed using captransf program (from Trusted-Logic ) and you should use a captransf compliant loader.

I loaded some applets using Axalto  /  Javacard SDK Tool running on Windows.  Contact Santiago Cabeza Sos  from Axalto for tool availability. NOTE: the tool is not tested on Linux, so it does not run on Linux!  Anyway, as it is a Java program using OCF, it should run.....

On Linux, I tested a simple applet with  pcsc_tools / scriptor.    Thanks to  Ludovic Rousseau [mailto:[EMAIL PROTECTED]]
No problem.

I hope it secures you about  Cyberflex Access cards on Linux.


Best Regards.
Philippe Smadja.

NOTE:    Linux egate-driver requires libusb 0.1.7 (KO with libusb 0.1.6-3).

At 04:47 AM 12/23/2003 -0500, Loic Prylli wrote:

Given the presence of some references to the "Cyberflex 32k" card in the musclecard software and in the mailing-list archive,
 I bought a couple of  them to use for various authentication use, but I found myself stuck :-(
My conclusion after som eresearch summarized here seems to be that the variant of the card they  sell now
 "cyberflex acccess 32k egate"  is impossible to use with musclecard or probably any other javacard applet,
(at least unless you buy their349$ SDK, but even then can anyone confirm that musclecard works with their SDK?).

The showstopper is that it is impossible to load the applet on their card with the following tool from Martin Buechler
(in theory this loader could work on other OpenPlatform javacard..):
http://vrweb.de/~martin.buechler/smartcard/CFlexAccess32Loader.zip (linked from http://www.musclecard.com/musclecard/index.html)

There are several references explaining this  problem. It appears to be caused by the factory enabled codeshield functionality
on their current cyberflex cards  (of course, you generally can only find that after losing a lot of hair). This functionality requires the applet
to have been preprocessed.

Note that before "codeshield" being the showstopper, you have to overcome a couple of easier problems described
in http://lists.musclecard.com/pipermail/muscle/2002-October/000225.html:
a) DEFAULT_CARD_MANAGER_AID should be A0 00 00 00 03 00 00
b) installing a propoer OCF environment

Then after this you would get the real problem (card returns error 69 85 => Conditions of use not satisfied) as mentioned in:
http://lists.musclecard.com/pipermail/muscle/2003-May/000785.html

The  codeshield technology which is suspected to be the cause of this error is described in:
http://www.cyberflex.com/News/Codeshield/codeshield.html or
http://lists.musclecard.com/pipermail/muscle/2003-May/000787.html
It is also somehow mentioned in the manual for the "cyberflex access programming" (available on http://www.cyberflex.com/Support/Documentation/documentation.html) which tells us that it is enabled
on the cards they ship now (search for codeshield).

*Ludovic Rousseau* mentioned in http://lists.musclecard.com/pipermail/muscle/2003-May/000828.html that it
might be possible to generate a modified applet using trusted logic software at:
http://www.trusted-logic.fr/download.html (select cap file transformer)
I have tried this. The available program (captransf) seems to terminate correctly after doing something on the musclecard applet,
but I have tried hard to load the resulting cap file, after putting it in the usual format (concatening the .cap components), but I still
get the same error.

To conclude this lengthy message, a few questions:
a) does anybody own the cyberflex SDK and could generate the .ijc file for the musclecard applet (and eventually send it to me)?
(according to the documentation ".ijc" is the suffix that the SDK should give to a post-processed "ready to load" applet)
Alternatively can anybody at least confirm that with the cyberflex SDK, musclecard works on recent cyberflex 32k egate cards?
(BTW: on a slightly different card somebody mentioned failure even with the Schlumberger SDK: http://lists.musclecard.com/pipermail/muscle/2003-May/000826.html).
Eventually could anybody with good knowledge about schlumberger ".ijc" files tell us if the captransf program
mentioned in the paragraph above should normally be able to generate acceptable applets, and/or if I miss something in my try?

b) Martin Buechler mentioned in http://lists.musclecard.com/pipermail/muscle/2003-May/000796.html
that he would like to remove his tool from the musclecard site. Would it be possible at least to include
a warning that this thing does not work with the currently sold cyberflex card, either on the website, or in
the Readme?

c) Finally, would anybody recommend a javacard that you can buy in a few units, and where you can load the musclecard applet?
 Otherwise what are currently the recommended (non-javacard eventually) cards that you can use with musclecard?
I haven't find a really up-to-date list on the website.

Best regards,

Loic


_______________________________________________
Muscle mailing list
[EMAIL PROTECTED]
http://lists.musclecard.com/mailman/listinfo/muscle

_______________________________________________________
Philippe Smadja                   R&D Project Manager.
axalto
A Schlumberger compagny.
36 rue de la Princesse, P.O. Box 45, 78431 Louveciennes C�dex, France
http://www.schlumberger.com
mailto:[EMAIL PROTECTED]
Phone: +33 1 30 08 48 54
Fax:     +33 1 30 08 45 24


Reply via email to