[comp.sys.apple] Duplicate mail

SEWALL@UCONNVM.BITNET.UUCP (04/22/87)

I seem to be getting two copies of much, but not all, of the
info-apple mailings.  Is anyone else getting multiple copies?
Possibly I'm on the list twice somehow?  I only sent a message
to info-apple-request one time - I hesitate to send a question
there as the result may be email in triplicate (oh bureaucracy!).

Is there any convenient way to "fix" the duplication problem
without either losing the mail altogether or getting 3 copies?

SEWALL@UCONNVM.BITNET (04/22/87)

I have heard from our moderator, and from several others.  Apparently
most receipients also are experiencing duplication.  Here's the gist of
our exchange.  Does anyone have an address of a network "guru" who
should be considering fixing whatever bug is at work here?

From:     Brint Cooper <abc@BRL.ARPA>

    You are not on the info-apple list twice; I just checked.

    There is no "bureaucracy" of list maintenance.  This is merely a
service provided on my own time and on a taxpayer-supported computer.
Use of a government computer is justified by the large number of Apple
computers at use in Army laboratories and offices.

    Duplicate messages often occur when a relaying host begins
"burping" for some reason.  Another cause of duplicate messages can be
found in the Internet/Usenet gateway where the newsgroup comp.sys.apple
and the mailing list Info-Apple interchange traffic.

 Brint (AKA Info-apple-request@brl.arpa)

My reply:

Thanks for checking, and for the prompt response.  I can survive
the deja vu experience.  I just received two mailings this morning
which are duplicates of messages from last Sunday.  Undoubtedly
the whole mail-list capacity over multiple networks is a complicated
software process.  If the unwanted duplicate problem exists on your
list, then it probably occurs for most others.  The total redundant
traffic on the nets must be substantial.  Given the concerns with
network message loads I've been reading about, it would seem a good
idea if someone who knows how the process works looked into fixing
it... I wonder who that might be?

I appreciate the info-apple service; it's very interesting.

ugcherk@sunybcs.uucp (Kevin Cherkauer) (04/25/87)

Just as a point of interest, I have noticed that no other board I have ever
read has so many problems with duplicate mail. I have *also* noticed that
no other board I have ever read has so many people on ARPA and BITNET
instead of USENET. Usually the Usenets predominate by a huge majority, but
here it's the other way around.
  There is probably a connection.

  Why is the distribution of users so different here? A (hazarded) guess:
maybe (just maybe -- no real proof or anything -- just a hunch) the reason
is that on other boards, people subscribe after already having an account
on a large machine for something else, and just happen to notice a newsgroup
that interests them. But here, most users are computer addicts, and have
scraped, clawed, and bitten *on purpose* in order to get access to the net
in any way they could.
  Does this make any sense? I know -- probably not.
  Any of you fit the scrape, claw, bite pattern out there?




--                                --  Kev  --                                 --
                 
..sunybcs!ugcherk                     "I *never* flame. *I* only flame *BACK*!"

SEWALL@UCONNVM.BITNET (09/29/87)

It appears that some gateways occassionally send mail to all the
"To" addresses on a received list even though some of the addressees
also are served through another gate.  How that can be fixed is beyond
me, but the appropriate list to post information about the problem is:

info-nets@think.com

Send along the complete headers (all that RECEIVED BY stuff) for a
couple of the duplicate messages so the net gurus can figure out which
combinations of gateways are duplicating messages.

Now that a single copy of info-apple is sent to NDSUVM1 for redistribution
to bitnet readers, we get far fewer duplicates BUT some still do show up
(apparently, some gate routes a second copy through the gateway at
wisc.edu).  Maybe watching for the appearance of duplicates at NDSUVM1
can identify the "abnormal" source?

---------------------
ARPA:   sewall%uconnvm.bitnet@wiscvm.wisc.edu       Murphy A. Sewall
BITNET: SEWALL@UCONNVM                          School of Business Admin.
UUCP:   ...ihnp4!psuvax1!UCONNVM.BITNET!SEWALL  University of Connecticut