Hello folks,
I wanted to point people to this draft RFC:
http://datatracker.ietf.org/doc/draft-kucherawy-dkim-lists/
by Murray S. Kucherawy, addressing DKIM issues related to mailing list managers. Murray has contacted Mark and I, and I'd like to invite your input on this mailing list.
-Barry
--On 14 May 2010 15:22:02 +0200 Barry Warsaw <barry@list.org> wrote:
Hello folks,
I wanted to point people to this draft RFC:
http://datatracker.ietf.org/doc/draft-kucherawy-dkim-lists/
by Murray S. Kucherawy, addressing DKIM issues related to mailing list managers. Murray has contacted Mark and I, and I'd like to invite your input on this mailing list.
-Barry
I'm curious about this paragraph: "Common modifications include: o Add header fields such as Reply-To:, Sender:, Resent-Sender: ([MAIL]), List-Id: ([LIST-ID]) or List-Unsubscribe: ([LIST-URLS]). In some cases, such header fields are replaced if the original message already contained them."
Does mailman replace such headers?
Is that desirable for umbrella lists? For example, if list B is subscribed to list A, then I guess the list-unsubscribe header does need to be replaced, since it's misleading to pass unsubscription details for list A to the subscriber of list B.
Is it still desirable for Mailman 3.0? I guess it'll be less common, since a single site will replace umbrella lists with rosters. However, if list A and list B are on different sites, the problem will persist.
What happens with digest messages? Is each digested message packaged sufficiently such that a DKIM message that it carries will not be broken? Is that possible? Does the idea of packaging single messages (like a digest of one) have any merit?
-- Ian Eiloart IT Services, University of Sussex 01273-873148 x3148 For new support requests, see http://www.sussex.ac.uk/its/help/
participants (2)
-
Barry Warsaw
-
Ian Eiloart