[Mailman-Users] Mailman is rewriting Message-Id

[Roel Wagenaar]

You wrote:

> Roel Wagenaar writes:
> 
>  > Seems to work fine, unfortunatily mailman is rewriting the Message-Id
of the
>  > injected emails.
> 
> Seems very unlikely to me.  The code you propose changing should prevent
> that already.  How did you determine that Mailman is rewriting the
> Message-Id?


The mails are coming from my main mailhost, collected via rss2email and
processed by procmail.
A copy of the local delivered mail is sent to the maillists server, after
delivering it loses its threading thanks to the newly inserted message-id.


> I suspect that what is actually happening is that you have an MUA that
> does not set Message-ID and you are CC'ing yourself to check.  Somehow the
> message gets to Mailman without a Message-ID, so Mailman (or something
> before Mailman in the process) adds one to the post it distributes, and
> something (probably your MTA on the way back in to your mailbox) adds one
> independently to the CC that you receive.
> 
> There are several other possible explanations (besides Mailman altering an
> existing Message-ID, which can't be completely ruled out but I'd bet a
> euro against a yen that something else is doing it), as well, but you
> haven't given enough information to diagnose the issue.
> 
>  > Is it save to alter /usr/lib/mailman/Mailman/Message.py so that the
section:
>  > 
>  >         if not self.has_key('message-id'):
>  >             self['Message-ID'] = Utils.unique_message_id(mlist)
>  > 
>  > is changed into:
>  > 
>  >         if not self.has_key('message-id'):
>  >             self['List-Message-ID'] = Utils.unique_message_id(mlist)
> 
> No, it's not safe.  Messages really should have a message ID according to
> the RFCs, and if Mailman receives a message without one, it should add
> one.  Lack of a Message ID will mess up threading in archives and MUAs, as
> most likely each downstream recipient will get a different Message ID,
> just as you do.
> 

Right now I am waiting for the next messages to come in and see wether  Mark
Sapiro is right with some headers being wrongly set.

Thanks so far.


-- 
Roel Wagenaar,

Linux-User #469851 with the Linux Counter; http://linuxcounter.net/

Antw.: Omdat het de volgorde verstoord waarin mensen tekst lezen.
Vraag: Waarom is top-posting een slechte gewoonte?
Antw.: Top-posting.
Vraag: Wat is het meest ergerlijke in e-mail?

A fire at one end, a fool at the other, and a bit of tobacco in between.