jon> Interesting. That wouldn't be too hard to write up... For my jon> purposes I just toss anything that gets held and I like that jon> setup, but I can see your point.
Well, if "just toss everything" means cranking through the web interface and clicking "discard" for each, that's the same amount of tediousness. If you mean just periodically ditching messages from the mailman queue behind the scenes, that's not an option for everybody. Also, don't forget that not everyone who admins a mailing list is also an admin for the machine (and maybe they don't even have shell access, etc). jon> Still the real solution here is to stop spam from entering your jon> site period. You should look at using the RBL and a few other jay> I agree with you on stopping spam prior to reaching the list(s), jay> but some times it takes a long time before the spammers IP Well, yes and no. It's a bit of "wrong tool for the job". In my particular case, I just want to make it easier to administer a members-only list. Mailman is very close to giving me just what I want. It knows exactly who the list members are without any guesswork at all. It's not clear to me that doing RBLs, blacklists, and other anti-spam stuff would be less hassle in the end. They're all heuristic in at least some minimal sense. If someone gets a bounce that says something to the effect of "members only, subscribe and then post", it's easy for them to fix that up themselves. (In fact, most of the time they may have just sent their posting from the wrong variation on their personal email address collection. Other times they just didn't realize the list was members-only; maybe they just browse the archives periodically to read it.) If someone gets a bounce that says something to the effect of "we think your ISP harbors spammers, so make them fix it or go away", that's a lot trickier for someone to get past. Some people don't have a different account/place from which to send mail. If that's so, they can't even send the admin a message asking to be approved on a one-off basis. That's an OK policy for many mailing lists (e.g., "Big Tom's Fly Fishing and Croquet Interest Group"), but for lots of other lists unfettered community input can be rather more important (ie, most IETF lists and the like). Anyhow, this isn't meant as a rant against various anti-spam measures. They certainly have their place and proven value in many contexts. RBLs, etc, are certainly the "real solution" for some problems, but not for this one. I'm just looking for the greater precision a Mailman solution could give. Thanks! -- [EMAIL PROTECTED] (WJCarpenter) PGP 0x91865119 38 95 1B 69 C9 C6 3D 25 73 46 32 04 69 D6 ED F3 ------------------------------------------------------ Mailman-Users maillist - [EMAIL PROTECTED] http://mail.python.org/mailman/listinfo/mailman-users
