Thanks for the responses to my original query about moving the Internet
addresses from LISTSERV to an Internet-connected Unix system.  Since the
responses have reminded me of several facts which I failed to mention in
my original query, here are a few clarifications:
 
(1) Yes, I know that running a mailing list with names stored in two
    different places would cause problems.  Hence my query to see if
    anyone had developed the tools to do it easily.  So far, no response
    that I've received suggests such a thing has been done.
 
    For those who say there is some intrinsic reason why keeping the
    names together is superior, let me point out that many mailing
    lists are divided between two or more nodes--we all know that this
    is S.O.P. within LISTSERV (which has the tools for doing it).
    At times in the past, any list with a lot of BITNET and Internet
    subscribers generally had a separate list on each side and this
    was considered the good thing to do in order to keep load off the
    gateway(s) between BITNET and the Internet and the links toward
    these gateways.  These days, the gateways and links are faster,
    there are more gateways, and Eric has worked to make LISTSERV
    create less work for the links, but I know there are still lists
    divided between Unix and VM nodes.
 
(2) Yes, my list (BIG-LAN) uses DISTRIBUTE (actually it uses DIST2).
    I should have mentioned this--it certainly is significant to the
    problem.
 
(3) Yes, I realize we have the software & interconnection to gateway
    my list to the Internet ourselves.  I would love to do this
    but am not allowed to.  We tried it in the past.  Suffice it to
    say that I have become convinced that I won't be able to change
    the minds of the powers that be.
 
(4) I thought my underlying problem (mail being routed to a less-than-
    perfect gateway) would be affecting other lists.  Just one person
    has responded to me saying he also experienced the same problem
    (with the same gateway).
 
(5) This problem has been consistently plaguing me for months, since
    early Fall.  I figured that any problems with DIST2's mechanisms
    would go away with a monthly iteration or two of the BITNET tables
    but in this case, that has not happened...and I'm still patiently
    discarding each error message I receive from such addresses so
    that when the problem IS fixed that these people will still get
    their mail.
 
John Wobus
Syracuse University