Re: Mailing lists - assumptions

On Sun, Apr 20, 2014 at 11:39:42AM +0200, Alessandro Vesely wrote:
> > The problem is that there might be some receiving servers that will
> > try to see if the originatingdomain.example.com.INVALID is a valid
> > domain address, and seeing that it isn't, might reject the e-mail on
> > that basis.
> 
> Exactly.  Why is that a problem?  Upon rejection the user will notice
> the ".INVALID" and remove it, just like he or she would have done if the
> rejection had arrived from dmarc-remediation.  The latter could have
> contained a pointer to an explanation page, which interested users may
> also find with google.

User "[email protected]" sends mail to a mailing list which rewrites his
from field to be "[email protected]".  When his post gets sent to
"[email protected]", hotmail.com could decide to bounce the
message because the from field is invalid.  And then just as in the
case of DMARC-induced lossage, this counts towards [email protected]'s
bounce count, and [email protected] could end up getting her
membership to the mailing list suspended.

If you change the from field to
"[email protected]", that's a valid
from field address, and reduces the chances that the receiving host
will bounce the message.

Part of the problem here is that mail can get bounced for many reasons
--- the user's inbox could be full, the recipient could be honoring a
DEMARC p=reject policy, it could just be the wrong time of month for
the mailer.  We could try to add some additional error codes that
indicate that the problem is not the recipient's inbox, but rather
that the incoming sender's from field, or SPF, or DKIM, or some other
arbitrary problem which is not specific to the recipient which is
causing the problem.  If we had that, then the mailing list could much
more easily decide that the right thing to do is to send the bounce
mail back to [email protected], instead of having potentially
autosuspending [email protected]'s mailing list service.  Of course,
then if [email protected] is subscribed to a mailign list with hundreds of
addresses that are all playing the DEMARK p=reject game, it might not
be pretty for [email protected]'s inbox....

						- Ted

_______________________________________________
ietf-822 mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/ietf-822