John C Klensin wrote:
So either put these clarifications into a separate draft and try
to advance it or start thinking carefully about what you are
wishing for.
+1.
I would suggest an I-D titled:
"Improving the reliability of Electronic Mail"
Hopefully more of a technical summary specification, stripped of
convoluted conflictive philosophies from mixed disciplines. At the
scale we are today, the Consensus (By Osmosis) IETF approach doesn't
seem to work too well - it partial explains all the deviant
implementations.
In all honesty, another rewrite of the "bible" - hosting requirements
guideline RFC 1113 would be a god-send.
The (rhetorical) question of the day would be, with RFC 2821bis alone,
would a new fresh engineer today be able to write an SMTP server or
client with the basic requirements all figured out?
For me, an opinion, 2821bis would be the end of line - a stripped
down, rewrite is most likely the best thing that can happen for future
generation SMTP developers.
--
Sincerely
Hector Santos, CTO
http://www.santronics.com
http://santronics.blogspot.com