At 9:16 AM -0400 4/5/01, CueMan is rumored to have typed:
(I guess that is to make sure the main list gets a copy of an email if someone mistakenly sends their post to the digest list instead of the main list)
That's one of the reasons I send any mail sent to the digest list's address to the interactive one instead. If I were using sendmail aliases (I ain't, but never mind that), they would look something like: list-l: "|exec /path/to/flist list-l" list-l-request: "|exec /path/to/flist list-l-request" list-d: "|exec /path/to/flist list-l" <--------- list-d-request: "|exec /path/to/flist list-d-request" Note that mail to list-d is piping to list-l. Clean, simple, and avoids mismailed problems entirely.* This _requires_ you to then use a recipe in rc.local.s20 to pipe a copy of the mail directly to flist list-d (since if you subscribe list-d to list-l, the post just gets sent back to list-l by sendmail). Of course, THIS is a good thing, since if you add a footer in rc.local.s20, you can make the pipe BEFORE you add it, eliminating all the nonsense people go through to _remove_ it in the digested list's rc.local.s00. (Gee, this is all much easier than I made it sound, honest.)
When the main list goes into moderated mode, the bounce copy needs the approval of the moderator and so is sent to the moderator for approval instead of dumping it.
Er...um...I don't see this behavior at all. Since the approved message is sent to the digest and then returned to the interactive WITH THE SAME messageID, the interactive doesn't send it _back_ to the moderators for re-approval, at least on my system. I send an approved message to the interactive, which pipes it directly to the digest; the digest _does_ return the message to the interactive (in my situation this is completely unnecessary but harmless), which eats the copy since it's "been here" before. Maybe you have a different problem? Is anything mucking around with messageID header fields? Charlie * Ok, ok, not entirely. There _is_ one anomoly with this...any -request email sent to the digest's list address will be processed by the interactive, not the digest. It's rare, but it does happen, and requires the listmaster to pay a little attention to misdirected unsubscribe requests. This is more a failure of the way SmartList kludges digested lists as completely seperate mailing lists, though, and to me is a completely acceptable problem for all the advantages it allows. --cfs3