This is a feature.  He's OKing the message from an address that is
slightly different from the address the OK confirmation was actually
sent to.  In his case it might simply be that the ID is differently-
cased.  I have this happen from time to time when I reply to a confirmation
request that was mailed to [log in to unmask] from my [log in to unmask]
mainframe address.  Since it's possible for [log in to unmask] and
[log in to unmask] to point to different mailboxes (although in my case
they do not), LISTSERV has to assume the worst-case scenario and responds
to both versions of the address.  You can alleviate this by ensuring
that the userid in the Editor= line is cased identically to the userid
produced by the editor's mail client.

This would also happen if the address the confirmation request was
sent to was mapped to a different address, from which the OK was
subsequently sent.  For instance if the Editor= address is
[log in to unmask] but the editor replies from his local mail
client which has a return address of [log in to unmask], this
would be the expected behaviour.  Again this is simple to fix, you
just make the appropriate change to the Editor= setting.

Nathan