From: Robert Klemme:
>
> "daz" schrieb:
> >
> > Robert Klemme wrote:
> > >
>
> [...] And of course we don't have a problem if mails reach the gateway in
> a differnt order than they reached the mail exploder, have we?
>

Definitely not.  In ML to NG direction, *all* mails should be mirrored
unconditionally to Usenet.  mail2news.rb has no checks for sequence.
A "X-rubymirror: yes" header is added so that when news2mail.rb reads it
from the news service, it will not be sent to ruby-talk ML.


> >
> > <from RFC 1036>
> > "The USENET News standard is more restrictive than the Internet standard,
> > placing additional requirements on each message and forbidding use of
> > certain Internet features. However, it should always be possible to use
> > a tool expecting an Internet message to process a news message."
> > </>
>
> Well, yes.  But that's just the message format, which seems indeed a sub
> format of SMTP.  But NNTP (and SMTP) is more than that.
>

I think I'm trying to make a distinction between Usenet and NNTP.
The important factor being that Usenet servers must pass messages
without modification of headers or content to the next subscribed
server.  The entry point for us is TU-Berlin's service.  The g/way
could send complete garbage out, but it should still be propagated
around the chain.  Of course, few news servers will accept it for
presentation to their users, so it won't appear.  Since I can
mirror a message that TU-Berlin "can't", proves that the message
is valid in the judgment of my news service and any that received
it later.
If someone is imposing censorship on Usenet, I can promise you that
there will be an "International Incident" which you will be able to
read about in Der Spiegel next week ;-))

Realistically, it's not going to be that.

>
> Now I don't even know who confused me.  Was it me? :-)))
>

At the time you wrote, you were considering the new problem
in the other direction as well and trying to reconcile the two.
I'm ignoring that totally, for now, to avoid confusion.
BTW: That post (your long reply to my long reply) was one that
didn't get to the ML.  So the new problem continues.  Hmm...


daz