On Fri, 19  Mar 1993 12:56:13 CST  Natalie Maynor <[log in to unmask]>
said:
 
>And you, as usual, sound like a  stuck record. (...) If we ditch bitnet,
>which I  wish we  would if  for no other  reason that  to make  you quit
>punching your boring send button and  saying the same shit over and over
>and over  again, then you  could no longer  say that anything  should be
>registered at whatever it is that your boring mail keeps going on and on
>about. (...) Bottom line: You're wrong.
 
Well I don't know about stuck records, and coming from you it is a pretty
ironical  accusation, but  I'll tell  you about  being wrong.  I've known
Melvin via  the network  for quite  a long time  and can't  remember many
instances  of him  being wrong  on technical  matters; in  fact, I  can't
remember any (unless you count the  usual "darn I hadn't seen that header
field, sorry  I just missed  the obvious"). But  then I'm not  a language
teacher so maybe I'm not qualified to judge :-)
 
As things  stand I'm glad I'm  sitting thousands of miles  away from you,
because  it looks  like you're  covered in  what you  claim Melvin  keeps
saying over and over  and over again. Since you do  not have a registered
gateway, you  are routed through  INTERBIT. By definition, INTERBIT  is a
generic service and  every single mail message routed to  INTERBIT may go
through  a different  gateway. Of  course they  will usually  tend to  go
through the same  gateway most of the  time, but the bottom  line is that
you rely on  a gatewaying service provided by volunteer  BITNET sites and
have absolutely  no control  over what  machine is  used to  gateway your
message. In other words, INTERBIT is not UGA. It was perfectly legitimate
for your mail to be routed  through Princeton, the delay was abnormal and
points to a  real problem somewhere between UGA and  Princeton or between
Princeton and your  system, which I would like to  see resolved anyway. I
would be grateful if  you would send Michael Gettes a copy  of one of the
messages which took 12-15h to reach you so he can investigate.
 
>The problem vanished as soon as  Harold fixed the UGA situation -- which
>is where the problem was.
 
That  was another  matter entirely.  Harold  wanted INTERBIT  mail to  be
routed through  his SMTP and  had inadvertently  caused it to  go through
another route. That was a problem in the sense that UGA wasn't doing what
Harold wanted  it to  do. It  was something Harold  could complain  to me
about. It wasn't a  problem as far as your delivery  is concerned - there
are sites which choose  not to route bulk mail to  their SMTP because the
load  is too  high and  NJE  deliveries cost  about 3  times less  system
resources. It  was not  something you  could demand  that Harold  do. The
problem you described remains unsolved, and I am not happy about that.
 
Finally I  would like to  add that the  PUCC gateway is  very competently
operated and that you would be wrong  to dismiss the problem as "heck, it
just means Princeton can't run a gateway".
 
  Eric