On 27/03/2021 21:00, Wolfgang Rosenauer via mailop wrote:
> Am 27.03.21 um 18:15 schrieb John Levine via mailop:
>> It appears that Wolfgang Rosenauer via mailop
>> <[email protected]> said:
>>> Yes, there are such cases but I don't think they apply for this list.
>>>
>>> Yes - the list breaks DKIM (which is already something which should be
>>> avoided since I do not see a need to modify the body with a footer
>>
>> It adds subject tags, too.  There are good reasons that lists modify
>> the messages.
> 
> Really better reasons than keeping authenticity measurements in place?
> I would say that's debatable.
> 
>>> But SPF passes in general for mails from that list because the sender is
>>> @mailop.org (not many lists do it like this though).
>>> Because of the above SPF is even "aligned" and therefore DMARC passes
>>> that message.
>>
>> Sorry, that's just wrong. For DMARC, SPF alignment means the MAIL FROM
>> domain matches the From domain.
> 
> And?
> From your mail:
> Return-Path: <[email protected]>
> From: John Levine via mailop <[email protected]>
> dmarc=pass (policy=none) header.from=mailop.org;
> because the mail was received from 2a03:4000:37:599:d8ce:dff:fee1:81c2
> which is permitted by the mailop.org SPF policy.
> 
> So what in my statement was "wrong"?
> 
>>> What I'm missing nevertheless as another mitigation on that list is an
>>> ARC-Message-Signature and an ARC-Seal from the listserver.
>>
>> Mailman 2 doesn't do ARC.  For that they need Mailman 3 or Sympa.
> 
> I didn't analyze why it's not in place but just stated that it isn't.

I have recently assisted FreeBSD mailing lists migration from Mailman 2
as Python 2 has become totally deprecated. We have decided to keep all
stuff for authentication and messages modifications inside Rspamd, well,
mainly because I can modify Rspamd to implement all features required
more or less easily. So far, we do all DKIM signing, ARC signing and
even DMARC munging within Rspamd only[1]. Therefore, any mailing list
solution can be used to serve ML traffic, even a most simple ones.

I have prepared a small presentation about that and several other things
[2] a couple of years ago (I'd like to say sorry in advance for my poor
English). Messages modification framework has made quite a significant
progress since that presentation.

[1]: https://github.com/rspamd/rspamd/issues/3647
[2]: https://papers.freebsd.org/2019/fosdem/stakhov-rspamd_freebsd/

_______________________________________________
mailop mailing list
[email protected]
https://list.mailop.org/listinfo/mailop

Reply via email to