John C Klensin wrote:
Pro: Clearly, a complete ABNF cleanup, to make the set closed
and consistent and, ideally, to eliminate and formalize as many
of the comments as possible would be a nice thing.
I'd class it as considerably more than "nice". The current document
uses ABNF at a level I would call "minimal" and I have occasionally
found some questions a challenge to resolve because of this.
Con: The odds of being able to do this in a reasonable amount of
time and to do so without introducing new errors --either within
Given that ABNF is, essentially, like writing code, any massive overhaul
to it would be certain to a) take a long time, and b) introduce new
errors. This is not the stuff of a bis or ter effort.
I suggest that new SMTP ABNF would make a nice, independent project for
someone. It would take the task out of the current critical path and
would permit the ABNF to be subject to extended and independent analysis.
If a community consensus emerges that the new ABNF is preferable to what
is already in the official SMTP, then it would be worth considering
re-issuing the document with those changes folded in. But what I've
described would take years to happen.