Stephen Paul Weber <singpolyma(_at_)singpolyma(_dot_)net> writes:
Is there any prior art on IETF specs having a "full" and "simple" form where
full implementations can read any output of simple ones, but not always vice-
versa?
Yes and no. I've tried this with the great pile of spaghetti that is the SSH
protocol (see https://www.cs.auckland.ac.nz/~pgut001/pubs/simplessh.txt), but
SSH has the same problem as PGP does, there's a single dominant implementation
that defines what the protocol is, and everyone else codes to that rather than
the spec (for SSH it's somewhat worse in that the client-side implementation
is Putty, which has bug-workarounds for all manner of broken implementations,
as a result of which it'll accept any old rubbish while at the same time being
used as a validation suite by developers of further broken implementations).
So even if you define a simple-PGP profile, I'm not sure how much (if any)
traction it'll get.
Peter.
_______________________________________________
openpgp mailing list
openpgp(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/openpgp