>   Can I  assume that if  one deletes  a user at  a BITNET node  (like a
>Canadian node that is  about to go off BITNET) - that  is to say someone
>signed up with a BITNET address - that it will also delete an entry with
>a domain (read it: INTERNET) style address that corresponds to it?
 
It is my understanding that this IS the "side effect" and is the reason
why the installation coded the :INTERNET. value to begin with.
 
Knowing what we know now, and assuming the NAD at the soon-to-be-free
of-bitnet-tryanny would be wise to start the UPDATE procedure one cycle
earlier (assuming that they do UPDATE at all) and simply remove the
:INTERNET. tag -- BUT that too will also present its own set of
problems especially for SEND=PRIVATE lists where subscribers have
subscribed under one network address and then e-mail from the other.
 
You're damned if you do, and damned if you don't.
 
It may be that you can "fake out" (bypass) the equivalence lookup
during DELete processing if you use some sort of wildcard for the part
to the right of the "@".  This you will need to discover for yourself
since I am not postive.
 
//Pete