Re: [exim4] Testing and making sense of smtp output
On Sun 19 Oct 2014 at 01:19:51 +0200, lee wrote:
> Brian <ad44@cityscape.co.uk> writes:
>
> >> > An address literal is not the same as an IP address. An MTA should not
> >> > be rejecting mail on the basis that the HELO is an address literal.
> >>
> >> Oh, then what is it?
> >
> > Using an example from RFC5321, an address literal is [123.255.37.2]. An
> > IP address would presumably be 123.255.37.2.
>
> Hm, there's not much of a difference, or is there? It's still an IP
> address and being used as one, only inside brackets for unknown reasons.
> Using IP literals when sending email used to work long ago ...
Earlier today I was using a dbus command and as part of it I typed
boolean;true
The command didn't work.
Looking closer at the web page I changed it to
boolean:true
Success! Not much of a difference surely; it's all punctuation. :)
> >> > It's probably academic what the HELO is most of the time. Many ISPs
> >> > will accept any old rubbish for it.
> >>
> >> That's a misconfiguration they should fix.
> >
> > You tell them. :) They might say they are not breaking any RFCs and will
> > accept any mail they feel like doing.
>
> I'm not sure if they aren't. The RFCs specify what the HELO must be
> like, and you could either argue that they comply with RFCs' policy that
> you should accept as much as possible or that you're breaking RFCs by
> using invalid HELO strings.
>
> At least they are supporting others in breaking RFCs, and I wonder how
> that could not be against their own interests. In any case, it
> classifies them as (at least potentially very) unreliable.
This is first time I've come across the concept of aiding and abetting
the breaking of RFCs. :)
(The idea that you are unreliable because you act differently is a very
dangerous one).
Reply to: