--On Monday, May 07, 2012 11:06 -0500 Mary Barnes
<mary.ietf.barnes@xxxxxxxxx> wrote:
> I agree with you John. I know that I have had messages
> discarded on several occasions. As a moderator, I look at the
> legitimacy of the posting and if it is a member of the
> community (it's generally very easy to tell), then I add them
> as being able to post even though they aren't subscribed. This
> can be very helpful as many of us use multiple addresses.
Yes. As a user of multiple addresses, one of the cases I was
worried about was precisely the one of someone posting from an
address that is not normally used with a particular list.
I just had a very helpful off-list discussion with Paul Hoffman
who pointed out that my suggestion was stated badly and easily
misunderstood. So, to clarify, assume that a message from a
previously-unknown address shows up on the list. Unless other
heuristics discard it as, e.g., malware, the message goes into a
moderator queue. There are then, I think, three possible cases:
(1) Moderator decides message is trash and discards it. As you
say, it is usually pretty easy to tell.
(2) Moderator decides the message is legitimate, allows it to be
posted, and authorizes that address to post to the list in the
future (as you suggest).
(3) Moderator decides the message is legitimate, allows it to be
posted, but does not authorize future unmoderated postings from
that address. This could happen for any of several reasons,
ranging from accidentally not getting the box checked to the
moderator's not being sure the message is really from an active
community member who is already subscribed to the list even
though the message content is fine.
For that third case, I was trying to suggest that it would be
good to automatically generate a message to the posting address
indicating that, while the posting had been approved by the
moderator, subscribing would be a really good idea if the person
intended to use the address in the future.
Paul pointed out that automatically generating such a message
would almost certainly require modifications to Mailman code and
that such modifications are a non=starter. That doesn't make
the idea a bad one, only one that is unimplementable with an
acceptable level of resource investment and possibly one that
should be passed on to the Mailman powers-that-be as a
suggestion.
I was not suggesting any change at all for the first case other
than we review our subscription and archive pages to be sure
they are adequately clear that postings from non-subscribers are
not guaranteed to appear on the relevant list.
best,
john