On  Fri, 2 Aug 1996 08:45:41 -0400, David B. O'Donnell - AOL wrote:
>In a message dated 96-08-01 18:04:02 EDT, [log in to unmask] (Rob Woiccak)
>writes:
>> just out of curosity, what constitutes a full mailbox for an aol
>>  subscriber?
>
>An AOL mailbox is "full" if there are 550 pieces of mail on the host system.
 
550 pieces is a LOT of email!  I've been a busy emailer for more than a
decade and rarely have a backlog exceeding 200 (of course, I do NOMAIL
lists when I leave for a vacation :-)
 
One way to get 550 pieces of email in a HURRY is to be the victim of one of
the bozos who subscribes people to a couple thousand lists.  Hence, it
*might* be useful to generate an UNSUB * (NETWIDE message to LISTSERV
_anytime_ a mailbox accumulates 550 unread messages (probably reducing
waste on AOL's end as well as ours).  Note however, that the netwide unsub
command ONLY works for LISTSERV.  Listproc, Majordomo, and other lists
would still generate round trip traffic.
 
Frankly, I find AOL a whole lot easier to live with than a number of
smaller hosts (that bungle their DNS entries regularly and can't always
remember from one day to the next who their users are).  AOL plays by the
rules, responds to inquiries, and generally pays attention.  Those of you
who remember the good old early, early LISTSERV days will recall that (it
seemed as if) every holiday in the UK was occasioned by some JANET mailer
daemon sending endless loop mail to multiple lists (mail queues of 10,000
overnight items per list affected at the US-UK gateway were not uncommon).
Only the fact that BITNET was operating at 9600 bps limited the number of
redundant messages received by list subscribers to a couple of hundred
before a gatemaster discovered the problem and purged the queue.
 
IMHO, compared to 10 years ago, list ownership these days is relatively
placid.  AOL is not likely to favor wasting resources bouncing mail to
subscribers that won't or can't read it.  Thoughtful, effective solutions
to identified problems might take a little time but are likely to be more
satisfactory to everyone than some quick fix that benefits only a limited
constituency.
 
For the impatient, there's a simple rule (I've mentioned it before).
Simply make all subscribers aware that your policy is: "If your mail
bounces for ANY reason whatsoever, you will be unsubscribed immediately.
Please resubcribe when the problem on your end has been addressed."
 
/s Murphy A. Sewall <[log in to unmask]>      (860) 486-2489 voice
   Professor of Marketing                          (860) 456-7725 fax
   http://mktg.sba.uconn.edu/MKT/Faculty/Sewall.html