Re: [Exim] Refuse connection if no MX for sending host

Top Page
Delete this message
Reply to this message
Author: Matthew Byng-Maddick
Date:  
To: Exim Users Mailing List
Subject: Re: [Exim] Refuse connection if no MX for sending host
On Fri, Oct 24, 2003 at 02:16:42PM -0400, Greg A. Woods wrote:
> [ On Thursday, October 23, 2003 at 18:58:22 (+0100), Matthew Byng-Maddick wrote: ]
> > Subject: Re: [Exim] Refuse connection if no MX for sending host

[The bit of your text you snipped follows:
>>> Yes, it is very evil, bad, wrong, and also a huge disservice to the
>>> community as a whole. Not supporting use of address literals is
>>> tantamount to not supporting <postmaster>, and in some cases it's
>>> exactly equivalent even.

]
> > Explain yourself.
> Why don't you read RFC 2821 Section 4.1.3 and learn for yourself?


Section 4.1.3 of RFC2821 starts:
|    Sometimes a host is not known to the domain name system and
|    communication (and, in particular, communication to report and repair
|    the error) is blocked.  To bypass this barrier a special literal form
|    of the address is allowed as an alternative to a domain name.  For


I don't think that answers your over the top assertion that not supporting
it is "a huge disservice to the community as a whole ... and in some cases
it's exactly equivalent...". It allows you to report some problems with MXs
that you can't do with RCPT TO:<postmaster>, however, if they are really
that ****ed with their DNS, chances are that you aren't going to know about
their mail (unless you read your logs as religiously as your inbox), because
it will have got rejected at SMTP time (and I know about the kind of policies
*you* apply, because we've discussed them before).

So, I'm still going to ask you why you think it's "equivalent" to not
accepting unqualified postmaster, because you've attempted, badly, to
fob me off. Explain yourself.

> > > Why the hell wouldn't you publish an MX for every mail server?!?!?!?!?
> > I think the question you wanted to ask would be better if it were correct:
> > ``Why the hell wouldn't you publish an MX for every valid domain-part?''
> No, not exactly -- I really did mean to ask exactly what I asked.


Oh, in which case you're stupider than I thought.

> What you are suggesting should be asked is very subtly different.


"subtly"? I think not. There's a huge glaring difference. So, I think it's
now up to you to answer the question you were asked.

I answered it, as did Tony, with "because it's not a valid domain part".

So, go on, enlighten us all, why *WOULD* you want to publish an MX for
every mail server? Please do it with appropriate (ie, quoted) references
to the RFCs that you are using to supporting your argument, rather than
like above pointing to an entire section that mention the four lines you
presumably mean, but also contain a full specification and grammar for
the implementation. (and another 7 lines of text following the four lines
I quoted, describing such in words).

MBM

--
Matthew Byng-Maddick         <mbm@???>           http://colondot.net/