When munging From and Reply-To, add original author To Cc instead of Reply-To

[Ralf Jung]

Steps to reproduce:
* Configure a list to munge both From and Reply-To (from_is_list and reply_to_list)

Actual behavior:
The emails will have a Reply-To field with both the list address and the author address. As a result, hitting "Reply" in a MUA (tested with Thunderbird) results in both of these addresses as recipients.

Expected behavior:
The mail should be set up in a way that the MUA behavior is like it would be without From munging: "Reply" has only the author as recipient, while "Reply All" has both the author and the list as recipient.

This can be achieved by adding the author to Cc instead of Reply-To in case reply_to_list is not 0. I attached a patch implementing this, that I tested successfully on my mailman installations.

This patch is against the CookHeaders.py of Mailman 2.1.18.

[Mark Sapiro (msapiro)]

Thanks for the report and patch.

There is still a situation that I need to consider further. If the post includes a Reply-To: with the poster's address and the list's first_strip_reply_to is No, then with your patch and a munged From:, the poster's address is in both the original From: and Reply-To: and ends up in both Cc: and Reply-To: in the delivered post.

I've waffled on this issue before, see <http://bazaar.launchpad.net/~mailman-coders/mailman/2.1/revision/1476>, so I need to give this some more thought. Additional input is welcome.

[Mark Sapiro (msapiro)]

My fix is somewhat different from the suggested patch. See <http://bazaar.launchpad.net/~mailman-coders/mailman/2.1/revision/1510>.

[Ralf Jung (ralfjung-e)]

This does not seem to work here. I applied the CookHeaders.py part of the patch on one of my systems, and for "Munge From", "Reply goes to list", the original author is completely lost. I couldn't yet figure out exactly what's going on - it seems the Cc header is prepared, but then not actually added.

[Ralf Jung (ralfjung-e)]

That should teach me not to apply half a patch... after adding your changes to WrapMessage.py, things are working as expected. I don't really understand what's going on, but oh well, the code is... a bit messy^^.
Thanks a lot :)

[Mark Sapiro (msapiro)]

Yes, it is confusing, but the general idea is that the changes to From:, Reply-To:, and/or Cc: that result from From: munging should be deferred until after the message passes through ToDigest, ToArchive and ToUsenet. Thus, the changes are actually applied by WrapMessage.

I will add some comments to that effect to the code to explain (and remind me) what we're doing.