[opensc-devel] [Muscle] secure pin verify/modify and byte ordering

[Ludovic Rousseau]

2009/3/21 Martin Preuss <aquamaniac at gmx.de>:
> Hi again,
>
> On Samstag, 21. März 2009, Martin Preuss wrote:
> [...]
>> The statement "Byte ordering is decided by machine architecture" is stupid
>> in itself: How would the reader know what architecture the device is
>> connected to? With this change the firmware is now supposed to somehow
>> acquire that knowledge in order to detect how the multi-byte values of the
>> CCID structures are to be interpreted...
> [...]
>
> Ah, sorry, I see: This was only about the PC/SC structures, not the CCID
> structures, right?

Exact.

You may also note that the fields in the PC/SC structure are not
exactly the same as in the CCID structure.
The PC/SC PIN_VERIFY has a bTimeOut2 field not present at the CCID
level for PC_to_RDR_Secure.

> Anyway, I don't think it was a good idea to change the statement about
> endianess, since now application *and* driver have to know what version of
> the PC/SC specs the other end uses...

Me too.

-- 
 Dr. Ludovic Rousseau