[Mailman-Users] From field displayed by MS Outlook

Barry Warsaw barry at python.org
Fri Jan 23 00:03:04 CET 2004

On Thu, 2004-01-22 at 16:48, Brad Knowles wrote:
> At 4:26 PM -0500 2004/01/22, Barry Warsaw wrote:
> >  Who's the agent responsible for the actual transmission of the message?
> >  Is Mailman acting as this agent?  Arguably not.
> 	I disagree.  I believe that this is exactly the role that Mailman 
> is fulfilling.

All I'm saying is that to me, the standards are unclear here.  Not
surprising, since this isn't the first time mailing list management
issues have exposed ambiguities in the RFCs.

The language in RFC 2822 seems to imply that Sender is the human agent
injecting the message into the network.  We're in the situation of
actually re-injecting the original message into the network, so would
the original agent be the Sender or Mailman?

This ties into the discussions we had earlier about munging Message-ID. 
Is Mailman within its rights to munge it, given that what comes into
Mailman and what goes out aren't exact copies, but these changes may not
meet the test of "subsequent revisions".

I actually don't think given the current slew of RFCs, that there's a
right answer.  Pragmatically there are probably not-so-bad and worse
answers though. :)

> 	However, you're not taking into account the fact that the MTAs 
> almost always ensure that there is a correct "Return-Path:" header, 
> and will in fact wipe out any "Return-Path:" header that might have 
> previously existed.
> 	However, this won't happen with the "Errors-To:" header, which is 
> largely used by older and obsolete MTAs which don't know about 
> "Return-Path:".

I hadn't thought about that.

> 	In this case, I disagree with the RFCs, at least so far as this 
> issue pertains to Mailman -- these rules are meant for MTAs and not 
> MLMs.
> 	It is highly unlikely to cause any harm for Mailman to use 
> "Errors-To:", and is likely to help when the messages are processed 
> by older and obsolete MTAs.  Moreover, it is useful to the rest of 
> us, because more modern MTAs will ignore it and yet it will still 
> provide potentially useful information.


More information about the Mailman-Users mailing list