wayne wrote:
[650 vs. 750 KB]
The actual survey was done around 2004/09/21, which is
September, not August.
This *may* be of some help trying to pin down how many SPF
records were published under the draft-mengwong-spf-0[01]
semantics.
Yes, but it's something the author of the SPF RfC has done
(i.e. you), citing the opposition (i.e. MS) might be a good
idea. Both estimates are older than senderid-core-00, the
best idea is probably to quote both.
There are also cases where a company has outsources some of
their email to an ESP and that ESP handles the bounces.
These emails will work under the SPF-classic semantics, but
not under the PRA without the ESPs and/or domain owners
changing stuff. This is the "Margaret Olson Objection", and
one that Harry Katz expressed some concerns about at the
Email Auth Summit.
Oh yes, I tend to forget that my "MSA 6.1 without 8.1" case is
a variant of the Olson objection, not exactly identical. OTOH
the potential "moderated newsgroup" issue is too esoteric for
an appeal, the chances are lousy that anybody at the IESG will
understand it (and besides there are other ways to submit an
article to the moderator)
Personally, I think the re-use of the Resent-* headers is
abusive. The PRA could just as easily have used a new header
since (almost?) no forwarders have ever used the Resent-*
headers.
I'm not sure about it, what did Dave really want in 822 ? It
is clear that it's "wrong" wrt 2822, but is it also wrong wrt
822 ? Where could it break existing software, if it's not one
of these obsure "header fields shuffled" cases ? In Bruce's
universe that's a real problem, but Bruce is no IESG member.
Bye, Frank