On Mar 12, 2015, at 11:01 AM, Werner Koch <wk(_at_)gnupg(_dot_)org> wrote:
On Thu, 12 Mar 2015 14:38, singpolyma(_at_)singpolyma(_dot_)net said:
- A v5 key format. Prepare for forthcoming public key algorithms.
What sorts of changes do you see being needed here?
In the past we collected several ideas for a v5 key format. We need to
revisit the list archives. IIRC, the plan was to first wait for the
outcome of the SHA-3 competition.
There were other things we talked about as well in V5. I think that connected
and separate, a hash-independent way to do a fingerprint is called for. There
was a nice proposal for that in the archives -- algorithm-id:hash-value -- the
abbreviated one.
that work is usually done under this WG as well. Do you see moving
more in that direction, or a full-on new RFC coming out of proposed
work?
Eventually a single new RFC should be done - for a v5 format this will
be needed anyway.
Or not -- it's a decision of the new working group.
We traditionally put everything into one document. That has a lot of
advantages. There's essentially one place to check. Other WGs -- for example,
S/MIME -- have a lot of separate documents. That means they can work on things
in parallel, but makes it harder to wind through the mass of documents.
I think more than one but less than fifty (which is only a slight exaggeration
-- the last time I checked there were over 35 S/MIME documents) is a good
compromise.
But yeah, I think you'd want to revise 4880 for new stuff. There are, however,
a number of ways to organize that.
Jon
_______________________________________________
openpgp mailing list
openpgp(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/openpgp