[Mailman-Users] spam, spamcop and mailman moderation
ge at linuxbox.org
Fri Nov 10 16:57:35 CET 2006
On Fri, 10 Nov 2006, Patrick Bogen wrote:
> I'm not entirely sure what the point of this message was.
> Bearing that in mind, you shouldn't be using moderation as a
> first-line anti-spam defense. Your MTA should be tagging emails as
> spam (e.g., using Spamassassin, or something better suited to your
> particular configuration), greylisting, etc. With a properly
> configured setup, the spam that actually reaches the moderation
> interface should be minimal; most of it should be discarded (not
> rejected) by mailman, at the very least.
> This is fairly trivial to implement; just set up your MTA to pass mail
> through spamassassin, and then add a check for the headers it adds to
> mailman's list configuration, if nothing else.
That fails the test of reality on lists I run which can be filtered. The
problem is so big now simple filtering doesn't do that much good. On those
lists that can't (security related with a lot of false positives) not
> If I'm understanding your concern, the key here is for you to
> configure your mailman installation to discard known spam messages
> rather than rejecting them. This is, in fact, one of the options on
> the moderation screen (you may choose to Accept, Defer, Reject, or
> Discard messages).
Auto-discarding may be an option, but it isn't in this case, as I need to
approve a lot of non-subscribed posts.
> Additionally, as far as I know, you CAN moderate non-members
> differently; although perhaps I don't have the same understanding of
> that phrase as you do. You can set messages from non-members to be
> automatically discarded or rejected, as you wish. See Privacy Options
> > Sender Filters > generic_nonmember_action
Moderate them differently and still have a choice? Going through modding
subscribers and seeing 2-3 posts, and going through non-subscribers and
seeing hundreds, simply isn't the same when on the same screen.
> - Patrick Bogen
More information about the Mailman-Users