Apologies if this is an FAQ, but I have checked the FAQ list and did some cursory scans of the (frighteningly voluminous!) list archives a while back.

Anyways, I've been wondering for some time why Mailman is designed around a once-per-minute queue run. Seems to me that this is the MTA's job -- and in fact, if things are going well (ie. if you follow the advice of FAQ #2 and turn off Exim's receiver_verify option, as I recently did ;-), then qrunner simply transfers Mailman's queue en masse to the MTA, which in turn will add problem messages to its queue.

So why the extra queue? Why not just pass messages off to the MTA as soon as they're received?

If this has been thrashed out before, just point me at the old thread (or, better, documentation) and I'll shut up ...

    Greg

-- Greg Ward - software developer gward@mems-exchange.org MEMS Exchange http://www.mems-exchange.org