From: Bruce Lilly (blilly@erols.com)
Date: Sat Mar 16 2002 - 14:30:33 CST
Henry Spencer wrote:
>
> On Fri, 15 Mar 2002, Bruce Lilly wrote:
[...]
> > Consider also mail-to-news gateways.
>
> Properly-implemented mail-to-news gateways have always had to work fairly
> hard to ensure that mail messages conform to the stricter requirements of
> news.
Assuming for the moment that there exists some reason to forbid
multiple msg-ids in Supersedes in Usenet articles, the question
of what a mail-to-news gateway should do when it receives a
message with multiple msg-ids in a Supersedes header remains.
As there is no defined relationship among the msg-ids in an RFC
2156 Supersedes header, there is no basis for selecting any one
msg-id for preservation. Simply deleting the header is probably
not the right thing to do, as that is likely to defeat the sender's
intent. The remaining options seem to be rather limited:
1. Inject multiple articles, each with one of the Supersedes msg-ids.
or
2. Reject the message.
I cannot think of any good reason not to permit multiple msg-ids
in a Supersedes header in a Usenet article. Some relaying and serving
agents might have to change slightly; the draft will require a great
deal of other changes, so that's not a very convincing argument.
Permitting multiple msg-ids avoids conflict with RFC 2156, simplifies
matters somewhat w.r.t. mail-to-news gateways and imposes no
requirements on readers, followup agents, injecting agents, posting
agents, or moderators. Optionally, one could specify an upper bound
on the number of msg-ids which a followup agent might place in a
Supersedes header (that limit might be set at one, leaving mail-to-
news gateways as the only source of multiple msg-ids, although 21
(same as for References) is probably more reasonable).