Re: [Exim] Domain literals: weighing up the arguments

Página superior
Eliminar este mensaje
Responder a este mensaje
Autor: Wakko Warner
Fecha:  
A: Exim User's Mailing List
Asunto: Re: [Exim] Domain literals: weighing up the arguments
> > > That's a good description of the way I have always understood this.
> >
> > No, it's totally bogus.
>
> So you're going to mandate that all hosts on the internet must accept
> email, whether they are running a SMTP daemon or not, just because
> you can mail postmaster@[theiripaddress] ?


IMO, Greg A. Woods is "totally bogus". What planet did he beam down from?

> This is the logical extension to the argument that postmaster@[ipliteral]
> works. Remember - one of the arguments put forward is that any host
> which can send mail should be contactable by this means so that problems
> can be reported.
>
> Consider the case where a host does not run a SMTP daemon, but has (eg)
> /usr/sbin/sendmail invoked by some user program which routes mail out
> from the host.
>
> Consider the case where a host sends mail out from behind a firewall,
> but does not allow incoming SMTP connections to that specific host.
>
> I'm sure there's many more cases which need consideration, and you'll
> soon realise that requiring postmaster@[ipliteral] actually only solves
> /one/ problem case.


I've wondered about this. Excluding what's supposed to be done and what
people do. Lets say that some "send only" host finds itself in some rbl. I
have myself (because my IP is labeled a dialup, but it's static) found my
machine in an RBL. Thus far it has been Wirehub/whatever it was, and Sorbs
(both DULs). It did not take long to remove myself from either. (Side
note, sorbs inherited wirehub's dul and their US system didn't know about
exceptions). Ok, lets suppose the only way to remove yourself is through
their web form and they will send to postmaster@[thatip]? What do you do?

> > You cannot look at this issue from one direction only.
>
> No, I think you're looking at the issue from one direction only. Take
> a moment to consider the side cases of wanting postmaster@[ipliteral]
> to work.


He's looking at the issue in the 'no' direction. =)

--
Lab tests show that use of micro$oft causes cancer in lab animals