If  I understand  it properly,  there  are two  main reasons  for the  current
proposed plan to split LISTSERV in  LISTSERV and LISTEARN: the assumption that
EARN will  migrate to  OSI soon,  and thus a  functional split  is unavoidable
anyway, and  Eric's desire of not  to be involved  any longer in any  way with
EARN regarding  LISTSERV. I share  Eric's concerns  and I can  understand them
perfectly,  but on  the other  hand I  think  such a  split would  be a  major
disaster, technically,  for the whole  network (not only for  EARN). Therefore
I'm  proposing an  alternative,  open to  discussion, to  try  to satisty  the
preceding reasons and  still delay as much as possible  the problems resulting
from the split.
 
I'd do the following: an EARN LISTSERV coordinator, paid by EARN, would be the
person responsible for LISTSERV in EARN. This person should be choosen to have
a very good experience  with LISTSERV, and should be the  final point for user
and  postmaster questions  and problems  about  LISTSERV in  EARN. Eric  would
forward unread  any mail  originating at  an EARN node  to this  person, maybe
sending at the  same time an automatic  mail to the originator  telling him to
direct further  questions to this EARN  choosen person. EARN would  pay Eric a
sum of money for the initial EARN-wide license for LISTSERV, AND a monthly fee
for the  privilege of obtaining some  maintenance and future versions  and the
burden of redirecting mail. The only person from EARN which Eric would have to
hear, for  the money  he would get  regularly, would be  this EARN  expert. He
might or  might not provide  fixes at the requirement  of EARN, but  anyway an
important bug would  normally also appear in BITNET or  NetNorth and would fix
things in  the next  release. Shipments would  be sent by  Eric to  BITNET and
NetNorth, and manually to  the EARN expert, who could decide  where in EARN to
install them, and thus should be the only one to contact, etc.
 
If EARN happened not to be happy with the level of service obtained from Eric,
or would need a  feature which Eric would not like to  provide (as for example
some OSI interface),  EARN would have the right to  continue using the present
version  of the  software, and  *at this  moment* the  LISTSERV/LISTEARN split
would be effected:  Eric would stop to receive the  monthly fee, the backbones
would be split, etc.
 
With this approach we would avoid major problems for a time, and Eric would be
free of EARN's nonsense.
 
  Jose Maria