On 23 May 2014, at 17:22, Tony Finch <dot(_at_)dotat(_dot_)at> wrote:
Sabahattin Gucukoglu <listsebby(_at_)me(_dot_)com> wrote:
Thanks for the extra Sendmail knowledge.
I was curious to see how my memory stacked up so I just grabbed the sources for
8.14.4. The option is still unset by default, and the generated cf files are
still using canonicalisation. This is all very silly. :(
The magic is all happening inside the DNS map, so all you really have to do to
turn this behaviour off is use FEATURE(`nocanonify’), which stops map lookups.
I hope that every Sendmail site does this. As a bonus your performance will
shoot through the roof, because you are turning off many DNS queries, even
those that do not result in CNAME expansion. The check is being done for every
RCPT command so if you want to send mailing list mail this is definitely
advised.
http://cr.yp.to/im/cname.html
Don't believe everything on that page. Sendmail does not do ANY queries,
and qmail's canonicalization code is bad in several ways.
http://fanf.livejournal.com/122220.html
He says: "If you set up www.your.site with a CNAME for your.site, mail to
www.your.site will automatically be accepted by your.site's mailer. Many
sites rely on this feature."
My mailers have never (well, going back to 1997 at least) done CNAME
canonicalization and this hasn't caused the interop problems you might
expect from many sites relying on this feature.
It didn’t make sense to me when I first learned about this particular “Feature”
of Sendmail, when I was first setting it up in 2004. Thankfully, before and
since, it still doesn’t. I well imagine that it’s just another piece of
prehistory that has got muddied over the years, but for my part I reject the
notion of application awareness of DNS, in this and other cases. CNAMEs are
just aliases and have no significance; in 2001 RFC 2821 was clear that this was
the case for SMTP and I configured my Sendmail with DontExpandCnames=True.
His comment about a 2001 target for deprecating the feature is almost as
amusing as the Sendmail documentation…
:)
Cheers,
Sabahattin
_______________________________________________
ietf-smtp mailing list
ietf-smtp(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf-smtp