After reading messages on this list it gets pretty clear that smart card middleware must be one of the most difficult things you can think of.
As some of you know, I'm in the (long-term...) process of defining a "smarter smart card". Eventually it should have its own USB-class but that is slow route which would disable the chance of making a proof-of-concept in a reasonable fashion. So (inspired by serveral other who have proposed using files rather than APIs), I have come up with the idea that the smarter smart card maybe could reuse the standard mass memory interface but only exhibit two predefined virtual files "input" and "output". I.e. each API call would be translated to a stream of bytes written to "input" while the result would be taken from "output". This scheme should also allow full 12 MHz operation which is rather much a requirement since a single Information Card may be as large as 100Kb. Naturally there must be a PKCS #11 as well but it will be extremely thin, what's not so thin though is the "air-tight" provisioning system but it can be written in Java and be 100% platform independent. The (anticipated) development system: http://mbed.org The core principles behind the smarter smart card: http://webpki.org/papers/keygen2/secure-key-store.pdf Comments? Anders _______________________________________________ opensc-devel mailing list opensc-devel@lists.opensc-project.org http://www.opensc-project.org/mailman/listinfo/opensc-devel
