pem-dev
[Top] [All Lists]

Re: summary of technical issues

1994-12-21 13:56:00
I sent my previous note prior to receiving this. At this point my 
primary concern is the impact the proposed standards will have on 
the evolving national/international public key infrastructure. I 
would prefer to go back to the standard X.509 certificate structure, 
using used-signed (self-CA) certificates as necessary until the full 
blown CA infrastructure can be deployed. 

There are valid and useful applications which operate outside of a "CA 
infastructure."  Self-signed certificated are not an interim step, they are an 
alternative certification model.  An X.509 is just a signed block of data 
which has a publick key and some attributes which the signer wishes to 
associate with that key.  That's it--any other interpretation is policy.  
X.509 certificates are just a representation for signing the combination of a 
public key and some attributes.

There will be multiple CA infastructures, and cases which operate outside any 
CA infastructure.  One of the major failings of Classic PEM (and the cause of 
a lot of the perennial discussion about what signatures "mean") is that it 
included a particular policy and certification model.  MIME/PEM, unlike 
Classic PEM, does not.  MIME/PEM does not depend on any national or 
international infastructure--it can just take advantage of one if such exists 
and is appropriate for a given context.

I think you're mixing policy and mechanism again.  MIME/PEM purposefully 
decouples key infastructure and policy from representation.
Amanda Walker
InterCon Systems Corporation

PGP Key fingerprint: 594F63C03B52DC4E37E9160DE733CD87
PEM MD5OfPublicKey:  8E4A21B7025943DE2EDC7CC038B3D6B1
<Prev in Thread] Current Thread [Next in Thread>