[conspire] Crossposts across multiple mailing lists (was: Even more FOSS events)

Rick Moen rick at linuxmafia.com
Fri Nov 5 13:43:45 PDT 2010

Date: Fri, 5 Nov 2010 13:42:15 -0700
From: Rick Moen <rick at linuxmafia.com>
To: Mark Terranova <mark at gidgetkitchen.org>
Subject: Crossposts across multiple mailing lists (was: Even more FOSS
Organization: If you lived here, you'd be $HOME already.

Oh dear.  Here we go again.

Quoting Mark Terranova (mark at gidgetkitchen.org):

> From: Mark Terranova <mark at gidgetkitchen.org>
> To: clswest at googlegroups.com, conspire at linuxmafia.com,
>         felton-lug at googlegroups.com, penlug-members at penlug.org
> Subject: [clswest] Even more FOSS events
> Reply-To: clswest at googlegroups.com

Mark --

I have no doubt you think the Reply-To you set fixed any problems that
otherwise would have been created by the crossposting.  Unfortunately,
that is not the case, because that is not how Reply-To works.

The Reply-To header is an RFC-mandated cue to MUAs and other
mail-handling software, by which a mail sender can indicate an alternate 
response address that the sender claims should be used for any
reply-to-sender responses.  So, by setting "Reply-To:
clswest at googlegroups.com", you are suggesting that any direct _offlist_ 
replies intended for you alone should go to clswest at googlegroups.com
rather than to mark at gidgetkitchen.org .

This is almost certainly not what you intended.  You were trying to say
'Yes, I'm crossposting this to four mailing lists, but any reply-to-all
response should be directed to clswest at googlegroups.com alone.

It's thoughtful of you to try that.  Alas, that is not what Reply-To
does, and it is not how it will generally be interpreted by the MUAs of
people hitting their MUAs' reply-to-all commands and responding to you.

The end result is that the various reply-to-all responses from up to
four mailing list communities are likely to cause considerable annoyance
to four mailing list administrators -- because it is very likely to
result in a great deal of held mail[1] from non-subscribed sending
addresses, on each mailing list to which a responder is not subscribed.

In the past, people who crosspost have created quite a lot of problems,
and for years I asked the main offenders politely to cease doing it.
There were several (I am of course _not_ speaking of you, here) who
consistently ignored my polite requests and explanations of why
crossposting is a very bad idea.  They seem to have treated all of
those discussions as my attempting to coax a personal favour out of
them.  In the end, I lost patience with asking for polite behaviour, and
inserted several anti-spam filters on several of the mailing lists I
administer, automatically intercepting and rejecting the most common
crosspost combinations.  I hated to do that, because I dislike
heavy-handed mailing list administration, but the chaos created by
crosspostings, and by the careless and inconsiderate people who blithely
cross-replied to them, kept being a big enough problem that I had to put
an end to it.

I've not yet created filters autorejecting crossposts between
conspire at linuxmafia.com and either the PenLUG, clswest, or felton-lug
mailng lists, and I'd rather not need to, but I absolutely will if this
keeps happening. 

You might ask:  If not crossposting, then what is better?  Fair
question.  The least problematic way I know of, of plastering announcements 
across multiple mailing lists, is to multipost, rather than crosspost.

(One of the worst crossposting offenders rejected that suggestion as
'too much work', and kept crossposting until I stopped asking for polite
behaviour and just installed reject filters.  Remarkably, he adopted
polite habits with zero complaints, the very instant the rude ones
ceased working.  There's a lesson in that, somewhere, and I fear that it
feeds my cynical side more than a bit.)

[1] _Almost_ all mailing lists in 2010 are configured to hold for
listadmin scrutiny postings from non-subscribed addresses, for antispam
reasons.  There are still some exceptions, but they are few indeed.

More information about the conspire mailing list