1. First since you're talking about promises made at the closure of MARID,
can I ask if you intend to follow the this:
http://www.imc.org/ietf-mxcomp/mail-archive/msg05054.html
"The Area Directors intend, therefore, to request that the experimental
proposals be reviewed by a focused technology directorate. This review
group has not yet been formed but, as with all directorates, its
membership will be publicly listed at
http://www.ietf.org/u/ietfchair/directorates.html
once it has been constituted."
I know the directorate called "DEA" has been formed, but last I checked
there was no mention about it on the directorates page above and no
information on who the people who are reviewing documents are and even
which documents they are reviewing.
I would be very interested in having this information be available now that
SID documents are being reviewed by IESG because I believe that correct
way to bring the issues with drafts to the attention of IESG is through
the people who are in this directorate. Without this option, I may have no
choice but to directly bring the technical issues directly to IESG which
considering IESG already busy schedule may not be what you want.
Note: You can safely assume that I've already tried to have the issues
brought up with draft authors and they are not inclined to even
mention it in the draft. And I do believe the issues are such that
would make it impossible for some of these "post-MARID" drafts to
ever become a standard in their current form and are likely
inconsistent with even an Experimental use situation.
I'm also very disappointed that there does not appear to be any correct
ietf public forum to review and discuss post-MARID drafts, nor is it clear
if there would be a last-call. While Wayne for SPF-Classic draft chose to
voluntarily bring his draft for reviews at ietf-822, ietf-smtp and
namedroppers, authors of other documents are unfortunately not doing so,
which is in my view may result is severely worse documents that have not
undergone appropriate standards review.
2. Now I'll comment giving my purely personal view as one of MARID
participant on this ongoing MARID-related confusion situation.
My comments are rather long, you may skip it if you like although
reading them may give you good view of the situation from the
perspective of somebody who is/was active at both MARID and SPF
On Mon, 23 May 2005, Ted Hardie wrote:
What follows below is, indeed, a demonstration of confusion,
but it is on Wayne's part. Wayne elides the fact that the
ASRG work had been going on for some time and that the
SPF work was understood in that context.
I agree with you.
Further, one of the original
SPF authors, Mark Lentczner, (cc:ed on this message but
not Wayne's) was one of those who requested a BoF and
then working group for MARID.
He was the only one who requested it, there was a lot of hope from many
people when that WG was formed and closing of the WG was certainly viewed
as disappointment considering it was just then beginning to make progress.
I do want to bring certain facts about how original draft was made (which is
also true for current drafts). Basically Mark was not as much an "author" as
he was what in IETF terms is called an "editor". The draft is/was something
constantly discussed at spf-discuss list and it is made up of various text
parts suggested by individual contributors and either directly used or
summarized by the draft maintainer, originally that was Meng, then it
become Mark and now Wayne is doing it. I realize that is basicly how IETF
WG work as well and sometimes its difficult to tell when draft is really
made by several "authors" or when listing as editor is more appropriate
and I imagine for individual submissions this is even more difficult as
you'd never know that certain submissions are result of group work.
Mark was the person who
originally produced the post-Marid SPF draft;
Wayne and the spf-council asked to replace it. Mark agreed on
the condition that they produce something reasonably
quickly.
After closure of MARID there was a lot of discussion if we should continue
working MARID branch of documents or should try to document the existing
and currently deployed "spf1" i.e. "SPF Classic" system. I'm sure you
can check spf-discuss archives, but I believe the position has been made
clear that almost everyone there wanted to get "SPF Classic" documented
and made into Experimental or Standard RFC as first priority with second
priority being further work on UnifiedSPF and post-MARID 2.0 specs with
support for multiple identities including PRA.
Further complicating things was that it was thought that we did not have
much time and original spf (spring 2004) drafts were not in as good a
shape as far as the text as marid protocol document (which has undergone
partial review through IETF WG). Mark said it would be easier to try to
adapt MARID protocol document to "spf1" rather then the other way around
and considering he was an editor, everyone agreed that he should do that
if he things it would best to get document published right away.
Personally I believe the result was rather disappointing and post-MARID
draft Mark published did not document original SPF well and also
inherited some problems from MARID protocol that still needed to be fixed.
(I dont blame Mark - there was really not enough time to produce anything
better). I suspect that Wayne felt the strongest about this and started
working on his own draft version which after he published was well liked
by people who are implementing SPF and so after formation of formal SPF
Community steering group, the draft was formally adapted as official
group document.
That document continues to see reviews at the mail list with discussions
about some aspects (such as result when unknown domain is in "-redirect")
and IETF related specification (like header fields registration) are now
all fixed as well. I'm amazed how much improvement there has really been
in this document from -00 to -01 even though we thought -00 was pretty
good draft back then.
This was agreed to by Wayne. At no time during
that exchange was a change in the function of that document
discussed or considered.
I believe it was misunderstood by many including both you and Mark and
others at SPF Community. We really should have discussed it and should
have continued working on MARID-chain of drafts that bring proper scoping
support in addition to documenting existing SPF1 draft. I personally was
hoping that spf-classic draft would be published quickly and within 2-3
months (from Oct/Nov 2004) we'd be working on UnifiedSPF that supports
multiple scopes (including MAILFROM, SUBMIT, PRA, HELO, PTR, etc) to
be documented by separate draft documents all going as EXPERIMENTAL.
I'm disappointed that it did not happen, its definitely quite possible
to work on it and work out proper scoping syntax within context of
existing spf1 records. This would obviously require equal participation
of all interested parties including Microsoft.
But as far as existing document, Wayne is right - its intent (as well
as that of Mark's post-MARID draft) as understood by people at spf-discuss
is to document existing ongoing use of SPF Classic (aka spf1) and not a
replacement for MARID documents. Its very unfortunately that this confusion
has only become obvious now after many months.
I certainly hope that resolution can be found quickly. Its possible that
the only way out (as not to break promises originally made) is to actually
start from the MARID-end again and to bring back Mark and ask him to work
on marid-protocol document as individual submission, hopefully incorporating
some improvements from spf-classic drafts (this would obviously not change
the fact that SPF Community wants to send spf-classic draft for IETF review
and publication as RFC).
Its also possible that you may be able to find other way to resolve this
even if it means it is not 100% as what was agreed post-MARID, I would
certainly hope that you working on behalf of IESG can work with SPF
Council to that end.
If the SPF council would like to publish something other than
the Experimental draft understood in the MARID post-morten,
they are welcome to issue whatever documents they like in
whatever document series they control. But the authors of
the SPF and Sender-ID drafts considered in MARID did agree to
Experimental publication
Unless I'm mistaken and misunderstood, SPF was never considered at
MARID, at least not "Classic SPF", i.e. SPF1. What was discussed at
MARID was using SPF protocol specification 2.0 with MAILFROM and SPF
protocol specification 2.0 with PRA.
That is however NOT what either draft-schlitt-spf-classic draft describe
NOR is it what draft-lyon-senderid describes. Both of them changed
semantics of the records and protocol used since the time of MARID.
I hope that my comments and views of the situation were helpful to IESG.
Regards,
--
William Leibzon
Elan Networks
william(_at_)elan(_dot_)net