[Mailman-Users] receive your own posts problem

Stephen J. Turnbull stephen at xemacs.org
Fri Apr 12 05:48:37 CEST 2013


William Bagwell writes:

 > So if a list adds a footer to the body of a message (many do) then that 
 > implies that the Message-ID /should/ be changed.

No.  As the section you quoted later shows, that is a "syntactic
difference" and clearly *not* a reason for changing the Message-ID.

 > Obviously Mailman does not want to break threading for no good
 > reason. However fighting Gmail's annoying feature with a less
 > annoying kludge is a good enough reason in my book.

I didn't say that individual site managers couldn't do this.  I think
it's a bad idea for Mailman to provide the feature.  I'll probably
lose, just like we've lost on Reply-To munging.  Note that it's a
10-line or so Handler, which you place near the CookHeaders Handler.

I realize that having Mailman provide the option is the most
convenient way to handle the issue, and if individual senders consider
that the message in their Sent folder, the copy that they receive via
self-cc, and the copy they receive via the list are different, then
they are arguably within the RFC to have the Message-IDs changed.  I
think that's a misinterpretation of the RFC, but I doubt even the RFC
authors have thought carefully about Gmail's perverse behavior in this
context.

Note that there are already ways for senders to handle the issue.  For
example, have Mailman ACK your posts, which tells you Mailman has
received and distributed the post.  Accept that sometimes you're
boring and nobody replies -- if they do, Gmail does display your sent
posts as part of the "conversation", which should be fine.

Of course, this is a little painful for the sender.  But that's as it
should be, since "We know you DO have a choice of airlines, and we
thank you for traveling with United today."  If you don't like the
inflight meal, pay for business class.  Get another MUA.

 > If the sender ticks the box telling Mailman to change the
 > Message-ID then they get to take the heat.

No, the sender gets exactly what she wants, and is very unlikely to
take any heat.  Mailman will get to take the heat when their non-list
correspondents complain, usually directly to list owners rather than
to senders, and occasionally to Mailman-Users.  You understand the
consequences of that choice, but in my experience most users won't,
and will impose problems on third parties without hesitation.

This feature, like Reply-To munging, is an invitation to smoking in
the elevators.

 > Other settings have brief descriptions / warnings so a sentence
 > warning about breaking threading

It doesn't break threading, except for the sender (that's exactly what
they want), and Cc/Bcc-only recipients (who don't deserve this).  In
fact, according to your argument claiming the feature conforms to RFC,
it doesn't break threading at all, since only the message distributed
via the list is part of the list thread.  The other message generated
by the proposed feature is in a separate subthread.

 > and the other issues you raise would not be out of place.

"If you enable this feature without complaining vigorously and
periodically to Gmail, then you are a selfish antisocial wretch" 
would do in my book. ;-)

An alternative to telling them they're being selfish would be to
automatically enable ACK on people who enable Message-ID munging to
remind them of the trouble they may be imposing on others, and give
them an incentive to stop it.

Yet another alternative would be to reject list posts that contain
CCs.  Then *all* problems are restricted to the sender and Bcc
recipients (but they're unusual as far as I know), because nobody else
ever sees the original Message-ID.


More information about the Mailman-Users mailing list