Subject: Re: Possible to build a powerpc disk (disklabel + fs) on an i386?
To: None <tech-kern@netbsd.org>
From: der Mouse <mouse@Rodents.Montreal.QC.CA>
List: tech-kern
Date: 11/20/2001 12:51:46
[This really doesn't belong on tech-kern.  Surely we can find a better
place for it?]

>> As I read it, the "verification" that MUST NOT is talking about is
>> the verification the previous sentence mentions, [...]

> That isn't what was intended, nor does there seem to be any rational
> justification for that

It doesn't make a great deal of sense to me either; the only reason I'm
doing as little checking as I am is that MUST NOT.

> The intent was that as long as the syntax of the command is OK, mail
> connections are not to be rejected.

That's the second time you've said what the intent is.  Were you
involved in writing it or something?  (ie, what is your basis for these
statements about the underlying intent?)

> The kind of behaviour that your server is (apparently) following
> required me to add a dummy MX record for my laptop, just so broken
> mail servers would allow it to use its name in the HELO (or EHLO)
> command.

If you don't have a stable name, you probably should be smarthosting
your mail to someone else.  But that aside, this sort of case is why
2821 permits bracketed dotted-quads as EHLO/HELO arguments.  2821
section 4.1.1.1 (discussing EHLO and HELO) says that

      ....  The argument field contains the fully-qualified domain name
   of the SMTP client if one is available.  In situations in which the
   SMTP client system does not have a meaningful domain name (e.g., when
   its address is dynamically allocated and no reverse mapping record is
   available), the client SHOULD send an address literal (see section
   4.1.3), optionally followed by information that will help to identify
   the client system.

2821 section 3.6 also says

   -  The domain name given in the EHLO command MUST BE either a primary
      host name (a domain name that resolves to an A RR) or, if the host
      has no name, an address literal as described in section 4.1.1.1.

Your MX isn't enough; you should send something that has an A, or an
address literal.

> Its address keeps changing,

Address literal.

> it often gets stuck behind NAT boxes (which it is now),

Yet another of the things NAT breaks.  If you can't even find out your
own address you arguably shouldn't be talking to the net-at-large.

> I don't expect it to ever receive connections from anywhere, so
> there's no point having it in the DNS.

I don't see why the one follows from the other.  If you don't accept
*or originate* connections, I might agree....

> Sure, I could have it send a fake HELO command, but isn't it better
> for it to send a real one?

Certainly.  Send a bracketed dotted-quad, an "address literal".

> 2821 says nothing at all in particular about this issue.

See above.  Also read the rest of section 3.6, which requires that
_any_ domain name used in SMTP be resolvable to an A or MX.

/~\ The ASCII				der Mouse
\ / Ribbon Campaign
 X  Against HTML	       mouse@rodents.montreal.qc.ca
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B