In article <v03130313b8ea0172f0a5(a)[192.168.123.10]>,
Charlie Summers <charlie(a)lofcom.com> writes:
> At 1:42 PM -0400 4/22/02, Tim Pierce is rumored to have typed:
>
>> This patch to rc.submit should make SmartList use the reject file to
>> screen list postings.
>
> Again, it seems a lot of unnecessary work, since (at least on every list I
> know about) if someone is consistantly sending unwanted posts, and refuses to
> stop after being asked to do so by the listmaster, that address is summarily
> unsubscribed. (And if you're determined to leave a specific address
> subscribed and ignore postings from them, a simple /dev/null drop in
> rc.local.s00 makes more sense than worrying about scoring from the reject
> file.)
We have found this feature useful for many reasons. Many list
managers want to have something between allowing people to post
whatever they want and unsubscribing them completely. For example,
they may be posting lists of lightbulb jokes and chain letters, or
they may be usually reasonable people who fly off the handle whenever
a particular person or subject comes up, or they may be carrying on a
temporary flamewar with one other person. While it may be reasonable
to unsubscribe the offender, depending on the severity of the offense,
some list managers prefer to leave them on the list but unable to
post.
I do not personally agree that a "simple" addition to rc.local.s00
makes more sense than using the reject file. There are several
reasons, but one of the most compelling was that I found our list
admins were already putting addresses of people they didn't want to
post into the reject file, then reporting it as a bug when their posts
were not rejected. They were showing me pretty clearly what the most
user-friendly implementation of this feature would be.
No one is required to make this change, of course, and I certainly do
not expect Charlie's views to agree with mine. :-) But, since the
subject of rejecting posts from individual subscribers, I am sharing
the solution that we found to work to this problem.
