"John" == John Levine <johnl(_at_)iecc(_dot_)com> writes:
John> The widely used fetchmail package includes a section of code
John> that determines the most likely responsible party for an
John> e-mail message. Its algorithm is similar to PRA but not
John> identical.
I think the fetchmail algorithm is a red herring. It basically boils
down to: determine the envelope sender from the headers. If you've
got a broken MTA, then use an algorithm which is similar to the PRA.
If that fails, there's yet another fallback.
If we're doing this at the SMTP level -- which is the prefered place
for the MARID check -- then the envelope sender is always known,
without any recourse to the headers, to this just boils down to using
the envelope sender (a la SPF).
Having an a complex fallback algorithm solely for use with post-SMTP
checkers that live behind broken MTAs seems the wrong solution for
MARID (though it was the correct engineering approach for fetchmail).
-roy