----- Original Message -----
From: "Koen Martens" <spf(_at_)metro(_dot_)cx>
To: <spf-discuss(_at_)v2(_dot_)listbox(_dot_)com>
Sent: Sunday, September 26, 2004 9:37 PM
Subject: Re: [spf-discuss] Re: draft-ietf-marid-protocol-03
BTW - this is exactly what M$ wanted - to mess up the opposition :-/
Look
at all the TXT records already published (some say half a million).
Are we
really going to tell all these wonderful people that they now have to
change
their TXT records into SPF records? What a mess :-(
Well, we might have to, since 'the DNS people' apparently are not happy
with our use of TXT records. Also, a dedicated RR is imho the preferred
solution.
In the long-term, absolutely. But we're in a kind of race to the patent
office here aren't we? So we need a working system quickly, and the TXT
record already is widely used and works.
It would be *much* better to come up with a method of accepting the TXT
records as they are now. M$'s new-fangled system might want other stuff
in
the TXT record, but surely we can filter that out at the MTA, and just
accept the SPF element of the TXT record?
But we also want new stuff in their for unified spf, and some other fixes
in the spf1 syntax (eg. looking up the domain tree if no spf records were
found). So we do need spf2.0 at some time.
Yea - I know, but v2.0 isn't out or working yet, and as I said, surely we
need a "quick fix" for now, while the long-term solution is thrashed out?
There's a long road of versions into the distant future, all with differing
demands of syntax and information which we can't guess yet, so while we need
the SPF RR to go forward in the fullness of time because it will deal with
those future versions, we also - and more importantly IMHO - need a solution
*now*.
It appears that SPF has lost a lot of time in the dance with M$, and we need
to catch up with where we would have been if we hadn't done that - if you
know what I mean.
Slainte,
JohnP.
johnp(_at_)idimo(_dot_)com
ICQ 313355492