On 12/18/14 11:06, Nikos Mavrogiannopoulos wrote:
On Wed, 2014-12-17 at 22:54 -0800, Jan Pechanec wrote:
+ Slot ID is a Cryptoki-assigned number that is not guaranteed stable
+ across PKCS#11 module initializations. However, slot description and
+ manufacturer ID may not be enough to uniquely identify a specific
+ reader. In situations where slot information is necessary use of
+ "slot-id" attribute may be justified if sufficient slot ID stability
+ is provided in the PKCS#11 provider itself or externaly.
Hello Jan,
I'd like to propose the following text instead:
"Slot ID is a Cryptoki-assigned number that is not guaranteed stable
across PKCS#11 module initializations. However, there are certain
libraries and modules which provide stable slot numbers and
descriptions. For these cases, when the manufacturer ID is not
sufficient to uniquely identify a specific reader, the slot
information could be used to increase the precision of the token
identification. In other scenarios, using the slot identifiers is
likely to cause usability issues."
That text discusses both the benefits and the risks.
If the consensus is that slot-id stays then I think text exactly like
you wrote is essential.
--
Darren J Moffat