>So the ability to handle this shortened/changed form of address might be
>good, but as I think about it, LISTSERV(tm) should (in the converse of the
>new 'probe' test) issue an 'Address has apparently changed' warning msg
>both to the user and the list-owner, so we'll have to deal with it anyway.
>It doesn't seem smart not to call attention to such changes and even if
>posting is allowed to pass through, the SET and SIGNOFF and other more
>serious account controls still need to be 'frozen' until the identity can
>be verified.
 
I love this idea. Let's have two different rejection notices - one sent when a
command or posting come from a totally unknown address, the other when the
address is "similar" (fuzzy match) to a subscriber's. Both
MAILTPL-cutomiwable, of course. This seems a more useful compromise than allow
posting/reject commands.
 
And zlso the CHANGE command, but that's pretty much orthogonal.
 
Nicolas