Thank you for useful advice. I will rewise the document, this is only
the first edition.
At 12.31 +0100 99-02-25, Charles Lindsey wrote:
In <v04020a02b2f9946b849d(_at_)[130(_dot_)237(_dot_)150(_dot_)138]> Jacob
Palme
<jpalme(_at_)dsv(_dot_)su(_dot_)se> writes:
4. Who may Post to a Mailing List
The checking on who may submit to a list is done on the SMTP sender of
the message, not on the names in From: or Sender: fields in the
heading.
What about messages arriving other than by SMTP (even uucp is still around
in some places)? In any case, if lists are nested, the SMTP address that
arrives may be from another list, whereas you may actually want to filter
according to the actual posters to that other list.
If you have nested lists, you usually have a top node,
where all new submissions should arrive. This top node
cannot let only subscribers post, because the subscribers
are on lists lower in the hierarchy.
I had not thought of this problem. A solution would be if
all the nested lists were members of each other, or if the
links in the nesting tree were doubledirected. Then anyone
could submit to any of the lists, and have the message sent
to all the other lists. Each list could then be set up to
check that only SMTP senders of its own subscribers
(including all the other lists) can send. This structure is
not so common, but would be a solution and should certainly
be described.
If you have a hierarchical graph without loops, then all
new submissions should send to the top node, so lists below
the top node should only accept submission whose SMTP
sender is the top node!
8. Loop Control
I would have thought the simplest means of detecting a loop would be to
add a Resent-Message-ID header containing a msg-id in some easily
recognisable form. Or else a recognizable Resent-From or Resent-Sender.
Is this commonly used? I thought the common idea was that mailing list
expanders should not add Resent- headers. Am I wrong?
10. Header Munging
Apart from what is specifed in sections 8. Loop Control and 9. List
Headers, a mailing list expander should not in any way modify the
heading of a message. In particular, the list should not change the
Message-ID, not add "Resent-", "From:", "Sender:", "Auto-Submitted:"
or "Reply-To:". The practice to add the e-mail address of the list in
a "Reply-To:" header is common, but is not recommended. Instead, use
the "List-Post:" command from [RFC2369].
I don't think you can deprecate the addition of Resent-* headers when they
are explicitly blessed by DRUMS as the proper way to do things.
For mailing lists?
------------------------------------------------------------------------
Jacob Palme <jpalme(_at_)dsv(_dot_)su(_dot_)se> (Stockholm University and KTH)
for more info see URL: http://www.dsv.su.se/~jpalme