[ On Tuesday, September 24, 2002 at 22:47:30 (+0100), Matthew Byng-Maddick wrote: ]
> Subject: Re: [Exim] Newbie SMTP/ISP-Problem ...
>
> On Tue, Sep 24, 2002 at 04:34:44PM -0500, Steve Drees wrote:
> > http://www.merit.edu/mail.archives/nanog/thismonth/msg00154.html
>
> I think you misunderstand. That mail is an argument *FOR* RFC2317, and
> against the idea that you should circumvent it in the way that URL describes.
Well, both Knowles and Pollard are overstating the issues to the N'th degree.
Pollard's scheme will work (i.e. Kowles is wrong), but as stated it also
does rely on a misfeature of BIND (so Knowles is not completely wrong),
though it need not do so if stated in a slightly different way. There
is no problem with a zone which contains only NS records and one PTR at
the origin of the zone.
Pollard's claim that the CNAME chasing required by RFC 2317 is too
complex and clever and contorted is just plain wrong -- the chasing is
required by the basic DNS standard RFCs anyway, so there's no reason not
to do things in the way recommended by RFC 2317. Just because Pollard
can't understand aliases doesn't mean the rest of us can't, and indeed
they can make things much clearer and easier to understand once one
accepts their basic concept. The authors of RFC 2317 certainly
understood the mechanism Pollard espouses, and may even have considered
it, though I don't recall if they ever did so in any published draft
version of the RFC.
I would say that because of the existance of RFC 2317, i.e. it's wide
publication and endorsement as a BCP by the IETF, Pollard's scheme is
the more complex one, though it is by no means contorted either since it
too is no different than any other subzone delegation.
--
Greg A. Woods
+1 416 218-0098; <g.a.woods@???>; <woods@???>
Planix, Inc. <woods@???>; VE3TCP; Secrets of the Weird <woods@???>
