ietf-smime
[Top] [All Lists]

Re(2): A draft ASN.1 module for Cryptographic Message Syntax

1997-11-20 10:41:43
But the cost of specifying and supporting two encoding options within
S/MIME (assigning parallel OIDs for each, etc) is quite significant,
and the result is ugly.

I'm not advocating that for all signature algorithms there should be
two object identifiers, one for the signature applied to CER encoding
and one for the signature applied to DER encoding. What I am advocating
is that CMS allows the choice between CER and DER when a signature
algorithm is specified.

For example, CMS might define one object identifier for RSA with SHA-1
on the DER encoding (for backward compatability), and another object
identifier for DSS with SHA-1 on the CER encoding (where presumably
we don't have a backward compatability issue).

Jim