ietf-smime
[Top] [All Lists]

RE: CERT-02 Comments

1998-03-24 06:09:46
Blake,

I need some explanation for this.  My understanding is that we have
three uses for certificates that would be affected by the keyUsage
extension:

1. Validating a signature on a certificate or CRL
2. Validating a signature on a message
3. Creating a RecipientInfo on a message (encrypting the content
encryption key)

Why are we calling out this specific case, but not the others?

My recommended text regarding the keyUsage encipherOnly and decipherOnly
bits defines how those bits should be used when the public key is being used
to form a pairwise key.  The X.509 and PKIX specs are not crystal clear
regarding that point.  Furthermore, I believe that the X.509 and PKIX specs
are clear regarding the use of the other keyUsage bits, so IMHO the S/MIME
cert spec doesn't need to say anything about those bits.

The X.509 spec, sec 12.2.2.3 states:

"h) encipherOnly: public key agreement key for use only in enciphering data
when used with keyAgreement bit also set (meaning with other key usage bit
set is undefined);

i) decipherOnly: public key agreement key for use only in deciphering data
when used with keyAgreement bit also set (meaning with other key usage bit
set is undefined);"


Applying the aforementioned statements to the generation of pairwise keys
yields the proposed text:

"If the keyUsage keyAgreement bit is set to 1 AND if the public key is to be
used to form a pairwise key to decrypt data, then the S/MIME agent MUST only
use the public key if the keyUsage encipherOnly bit is set to 0.  If the
keyUsage keyAgreement bit is set to 1 AND if the key is to be used to form a
pairwise key to encrypt data, then the S/MIME agent MUST only use the public
key if the keyUsage decipherOnly bit is set to 0."

If the group decides that the X.509 and PKIX specs are sufficiently clear
and that the recommended this text is not needed, that is OK with me. 


14) Appendix D, Please delete this entire Appendix because it is out
of
date
and not needed.

I will agree that it is out of date, but is it truly not needed?  Would
it be better to fix it or throw it out?

I recommend throwing it out.  The S/MIME v2 specs are readily available for
a side-by-side comparison with the v3 specs.


<Prev in Thread] Current Thread [Next in Thread>