What's wrong with PEM?

[Scott Collins]

After reading the RFCs for PEM (1421-1424), I am curious what other people
think about PEM.  For cypherpunks agenda, in what ways is PEM lacking?  My
take is:

  1. PEM is a protocol, only applicable to mail (perhaps only to internet
mail) while PGP is program that provides similar services for mail, but is
also applicable to non-mail related encryption tasks.

  2. PEM and PGP don't aggree on the symmetric algorithms (DES, IDEA).

  3. PEM certificates are bulky, and transmission is encouraged.

  4. PEM certificates are issued by Certificate Authorities, which would
seem to preclude PGP's 'web of trust' model.


These all seem to have answers:

  1. PEM is protocal, PGP is a program that implements much of what PEM
is... why not make PGP PEM compliant.

  2. Propose IDEA as a symmetric algorithm for PEM.

  3. Ha! PGP already has key servers.

  4. Propose a revion to the certification scheme where USER certificates
would be created by the owner and signed by non-certificate-authority
acquaintances ala PGP.


Yes, this would take time and effort.  No, this should not be taken as an
affront to our current and previous efforts.  I think that we should persue
_every_ avenue.  If the only real problem with PEM is the trust model, and
we can change that, then this would be a strongly legitimizing action.


Scott Collins         | "Few people realize what tremendous power there
                      |  is in one of these things."     -- Willy Wonka
......................|................................................
BUSINESS.   voice:408.862.0540  fax:974.6094   collins at newton.apple.com
Apple Computer, Inc.   1 Infinite Loop, MS 301-2C   Cupertino, CA 95014
.......................................................................
PERSONAL.   voice/fax:408.257.1746    1024:669687   catalyst at netcom.com