ietf-mxcomp
[Top] [All Lists]

SPF-Classic

2004-07-10 10:49:11

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

I've been following these discussions for some time and not had a thing
~ to say as it has appeared to me that a lot of balls are being juggled
simultaneously. That said I have 3 points I would like to make.

1. SPF Classic DNS records are dead simple to publish. This makes
adoption a matter of a few seconds per domain even for a DNS admin who
does his zone management on the command line. Ease of adoption is a huge
factor in the success of any new protocol or standard. I did all of my
domains (30 or so) in less than a half hour a few months back and it
just worked and it has ever since. It was simple to do and that ease of
implementation was a big factor in my taking the time to adopt
SPF-Classic. In the real world network admins are busy and do not want
to take time for "yet another major roll out" of some new high falutin'
technology (yes I'm southern born and bred so just deal with it! :-).

2. SPF Classic for the MTA is also ridiculously simple to implement and
this is another big boost as we not only want the DNS records published
but we need the MTA's to play their part as well. Now I'll grant that I
use postfix so that's the only MTA I've set up with SPF but it truly was
a simple thing to do. From what I gather the other OSS MTA's are as
simple or a bit more difficult but doable in any event. I have also seen
a commercial add on implementation of SPF for Exchange so it seems that
Microsoft's MTA is taken care of as well. This takes care of both sides
of the equation!

3. Given these two previous points what in the world are we doing
talking about changing things up? We have a working implementation that
fulfills the intent we all need. Why not write it up as is and have done
with it? The point of it all is that people adopt it yes? If they don't
we fail in our ultimate objective right? So why not go for SPF Classic?
It's proven and it works *TODAY*! Given that we have a working, free and
open source implementations available that is doing the job today for
thousands of domains already what is the reason for continued debate and
constant proposals for change? It seems a waste of time to me and as I
mentioned in point #1 above network admins are busy. It seems to me that
what's really happening is a bit of "jockeying" around for position and
advantage by a certain large, Washington state based corporation. Now if
I have mis-read the situation I apologize in advance but if my
suspicions are correct this is uncalled for and unwanted. The problem
we're trying to solve is immediate and growing... to overcomplicate it
in pursuit of some mystery corporate objectives seems to be a waste of
everyone's time but theirs. Is there some reason that SPF classic DNS
TXT records won't work with the Microsoft nameservers? I run bind on
Linux so I have no personal knowledge of it but I suspect SPF Classic
records ought to work just fine. Lets roll with Meng and have done with
the brouhaha!

- --
csm(_at_)moongroup(_dot_)com, head geek
http://moongroup.com
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFA8CwWv6Gjsf2pQ0oRAmtqAJ9MQ9+8JFe+6WXckiHUn+/rYQAnHgCgtFgA
LvZxw5SbHzUhvqHh4yWtOjs=
=7zP7
-----END PGP SIGNATURE-----


<Prev in Thread] Current Thread [Next in Thread>
  • SPF-Classic, Chuck Mead <=