[Mailman-Users] Reply-To: munging (was Re: [Mailman-Developers] Feedback ...)

Barry A. Warsaw barry at zope.com
Fri Mar 15 19:39:10 CET 2002

>>>>> "CVR" == Chuq Von Rospach <chuqui at plaidworks.com> writes:

    CVR> The other thing I'd like like to see is an option to 'strip
    CVR> all reply-tos' so that a reply-to that is set as it enters
    CVR> the mailman system is removed and not propogated. Allowing
    CVR> and end-user's reply-to to propogate can cause all sorts of
    CVR> havoc, and frankly, can be used to mailbomb someone if you
    CVR> want to play troll.

Let me summarize what the system does now (apologies if I'm just
repeating myself ;).  I think it is flexible enough to support any
Reply-To: policy your list owners might want to impose.

- You can config MM to strip any Reply-To: header on the original
  message, or preserve it.  This is orthogonal to any Reply-To:
  munging that may happen later.

- You can config MM to set a Reply-To: to point back to the list, or
  to any explicit address you may want.

- All the Reply-To: addresses that will show up on the reflected
  message are put in a single Reply-To: header, so as to be maximally
  RFC compliant (RFC 2822 doesn't allow for multiple Reply-To:
  headers, but does allow for multiple addresses on a single Reply-To:
  -- go figure).

    CVR> Bob's got a point. There has to be a way to say "don't
    CVR> override", although, if you don't enable the option for a
    CVR> list, don't you already get that?

I've totally lost context on this question so I can't answer it.

More information about the Mailman-Users mailing list