[This is a resend of my previous message with corrected addresses for Jim
Galvin and the ietf-open-pgp working group. Please discard my previous
Thanks for your kind words about MOSS.
Let me offer a bit of perspectrive about MOSS in relation to S/MIME. MOSS
was designed to take full advantage of MIME. It uses the MIME facilities
for hnadling multiple parts and for encoding binary information. Although
MIME is enormously successful, it does have some technical wrinkles that
complicated the design of MOSS and make the resulting design less elegant
than we would have liked. I'm referring specifically to MIME's requirement
that everything be encoded in a 7bit format and that the encoding be only
in the atomic parts of the structure.
Meanwhile, S/MIME adherents usually take the position that MOSS cannot be
implemented easily unless one has a complete MIME implementation already.
S/MIME, on the other hand, is designed to make maximum use of PKCS formats.
It implements enough MIME to be compatible with MIME format, but doesn't
use MIME mechanisms internally.
Let me also comment on one technical detail. You suggest that messages in
the MOSS forat could be transliterated easily into other formats. While it
is true that the text and structure of a MOSS message could be
transliterated into other formats, signatures are specific to the precise
encoding of a message and cannot be transliterated.
At the time MOSS was designed, PGP was not compatible with MIME and there
was no clear path for bringing PGP into the standards arena. That's since