[License-review] Approval: Server Side Public License, Version 2 (SSPL v2)

Rick Moen rick at linuxmafia.com
Sun Dec 2 03:07:15 UTC 2018


Quoting Florian Weimer (fw at deneb.enyo.de):

> * Stefano Zacchiroli:
> 
> > On Thu, Nov 22, 2018 at 09:18:18AM -0500, Brendan Hickey wrote:
> >> Replying here as Elliot's message seems to have vanished into the ether,
> >> despite appearing in the archives.
> >
> > FWIW I've had, and still have, the same problem with all of Elliot's
> > messages in this thread: they appear in the archive, but they do not
> > reach my mail server, contrary to the messages of all other participants
> > in the thread.
> 
> This is likely related to the DMARC policy published by the sender.
> 
> You need to contact your mail provider and ask them to override the
> DMARC check for mailing list email for your account.

Not the best solution.


The problem has two pieces.

1 of 2: Elliot's domain has a strongly asserted DMARC policy ('p=reject'):

Liten-Datamaskin:~ rick$ dig -t txt _dmarc.mongodb.com +short
mongodb.com.hosted.dmarc-report.com.
"v=DMARC1; p=reject; rua=mailto:1eed4417 at mxtoolbox.dmarc-report.com,mailto:dmarc_agg at vali.email,mailto:dmarc_reports at mongodb.com; ruf=mailto:1eed4417 at forensics.dmarc-report.com,mailto:dmarc_reports at mongodb.com"
Liten-Datamaskin:~ rick$

2 of 2:  ...and Stefano elects to receive his subscription copies at
GMail, which complies with (enforces) sending domains' DMARC policies
for all arriving inbound mail.  Hence, when the license-review MLM
(mailing list manager) redirector sends Stefano a copy of Elliot's mail
with rewritten headers, upon arrival GMail finds that it no longer
complies with mongodb.com's DMARC policy, and as requested rejects it as
a forgery.

DMARC (when fully enforced, as GMail does) is causing headaches for MLM
software all over the planet, resulting in refused deliveries and
(consequent) high bounce scores and subscriber autoremovals whenever
those two conditions occur.  Fortunately, recent versions of Mailman
support an effective mitigation (via header kludging) that sidesteps the
problem.  I've written offlist to OSI's listadmins to advise how to
enable that mitigation.

(Why, yes, I do administer Mailman in a bunch of places.  Thanks
for noticing.)

-- 
Cheers,                                         I could maybe do one pilate.
Rick Moen                                       -- Matt Watson (@biorhythmist)
rick at linuxmafia.com
McQ! (4x80)



More information about the License-review mailing list