Friends -
I have completed a draft that describes SPF v1 for submission as an
experimental RFC. I will do so one week from today. Please review it
and let me know any errors in it.
Let me remind you that the intention of this draft is to codify the
common understanding, implementation and deployment of SPF v1 (or "SPF
Classic"). It is understood that it doesn't capture changes
contemplated, implemented by some, or even already agreed upon for a
future version. Its main purpose is to have a published point of
departure for experimental deployment and future development.
With regard to my prior posts, I have answered my questions thus:
1) HELO domain checking has not been included. While it may make
logical sense, I read the postings as clearly indicating that its
semantics, implementation, and even its existence hardly constitutes a
common understanding. Note that the null reverse path rule, which uses
the HELO domain, is still included.
2) The Received-SPF header has not been included. Replies to my
question indicated general acceptance that this was part of SPF v1.
However, upon review of the available language, I have found that the
header is poorly specified: The grammar was incomplete and ambiguous,
and the operational aspects of it left some large open questions.
Hence, rather than design-on-the-fly, I left it out.
3) The new RR type language has been kept. Almost all replies were in
favor of keeping it and it seems clear that everyone understands what
its purpose and intent are, even if it is not implemented anywhere
(since, after all, we don't have a IANA assigned number for it yet!)
Therefore, without further delay, here are URLs to the drafts:
http://www.ozonehouse.com/mark/spf/draft-lentczner-spf-00pre1.html
http://www.ozonehouse.com/mark/spf/draft-lentczner-spf-00pre1.txt
- Mark
Mark Lentczner
http://www.ozonehouse.com/mark/
markl(_at_)glyphic(_dot_)com