If you add a non-zero setting for the

  * Auto-delete= Yes,Semi-Auto,Delay(4),Max(100),Probe(10)

Probe keyword, the server tends to "notice" failing addresses sooner in some situations.  It appears that some delivery errors are considered "temporary" unless/until they happen to a probe, and so might go on indefinitely if "passive probing" is not activated.  My experience is that if the passive probe message bounces, then LISTSERV starts using active probe messages at 0600 on consecutive days (regardless of actual traffic on the related list) until either the failures clear up or enough of them occur to cause automatic unsubscribe (in this example, four days).  That is, for immediate failures - things get more interesting when the failure is due to messages expiring on the SMTP retry-timeout for unreachable or just plain flaked-out sites.

When we have a set of subscriber addresses from some other source, to load into a new mailing list, I set the Auto-Delete=Probe value to 1 or 2, so that when traffic does start up on that list it will find the bad ones fairly soon, and once the list has 'aged' some move it up to around 10 or so (depends on how many subscribers).

>>> [log in to unmask] 06/11/01 03:51PM >>>
I'm not clear on how autodelete works: