ietf-mailsig
[Top] [All Lists]

Re: wildcards, was dkim technology?

2005-07-15 08:10:46

In <20050715070108(_dot_)25790(_dot_)qmail(_at_)xuxa(_dot_)iecc(_dot_)com> John 
Levine <johnl(_at_)iecc(_dot_)com> writes:

I suspect that the DNS gurus will give us (people interested in DKIM)
the same response as they gave SPF folks:  You can create a server
side macro processing that automatically adds the appropriate records
to the zone file before it is published.  The secondary NS will then
be able to pick up these extra records when they do an AXFR.

It would be useful to review the discussion of CSV wildcards on the
namedroppers list, particularly since DKIM uses name prefixes like
CSV does.

I think it would be more useful to simply get the namedropper folks
involved with discussions about the current DKIM proposal.


I confess that I don't follow namedroppers in detail, the only recent
discussion that I can recall of doing tree walks and CSV was in the
context of an SPF I-D review.  As such, it was somewhat off topic
because SPF has never proposed tree walks, so I suspect the depth of
dislike for tree walks was not fully revealed.


During the MARID interim meeting, Ed Lewis (DNS guru) made a
presentation about the "right way to extend DNS".  Tree walks were
explicitly given as A Bad Idea.  Other Bad Ideas were using TXT
records, using domain prefixes, and using domain suffixes.

Obviously, I don't agree with everything Ed said.


[...], and although a tree walk is a bad solution, [...]  they are
the least bad option.

One of the reasons why I proposed zone cuts for SPF a year and a half
ago was I was under the impression that tree walks are considered so
bad that they are unacceptable.

Again, I think that if we are planning on getting a spec accepted by
the IETF, we should probably wander over to namedroppers and raise
this issue now.  And, by "we", I think it would be best if one or more
of the major authors of the I-Ds does it, otherwise we might not get
as serious a review as we need.


-wayne


<Prev in Thread] Current Thread [Next in Thread>