Ian G <iang(_at_)systemics(_dot_)com> writes:
That's the sort of information we need. Is that sufficient to create the
paramaters and specify them in the standard?
I don't really know how necessary this would be in practice. S/MIME (which
Jon referred to in his original message) has a couple of algorithms in there
that were added due to various national requirements where the understanding
is that those who feel compelled to support the Wombat128 algorithm will know
the details of Wombat128 and don't need much in the RFC beyond an ID. Because
of this the S/MIME approach is to have a distinct RFC for each NIH algorithm
that copies 10 pages of boilerplate from the previous NIH algorithm RFC and
then changes a few sentences to say "Use this ID". I think this would be the
best approach for OpenPGP as well. We're well into the political layer here,
not much to do with technology any more.
Peter.