I support this draft.

I've read it many times in the past, and have implemented almost all
of it (no CRMF), but, as often happens when re-reading an old
document, I found a new issue, which could be construed as a nit if
the authors agree that my rephrasing is what they meant.

Section 3.9.7:

   In this profile a single reference object to publication location
   of the immediate superior certificate MUST be used, except in the
   case where a CA distributes its public key in the form of a
   "self-signed" certificate, in which case the AIA field SHOULD be
   omitted.

I think we need to change "MUST be used" to "MUST be present".  "MUST
be used" could be construed as constraining relying party behavior,
which would rule out mechanisms such as Steve Kent's algorithm for
constructing a local trust anchor.  Since the choice of a trust anchor
is, ultimately, up to the relying party, not the issuer, I don't think
it's reasonable for the profile to constrain the relying party in this
way.  So it's ok to require the issuer to supply the AIA, but not to
require the relying party to use it.
_______________________________________________
sidr mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/sidr

Reply via email to