[log in to unmask] wrote:
 
>Sorry but I've only just picked up this thread, and as a list maintainer
>myself I too am interested in finding a solution.
 
Well, that's encouraging, though one still gets the sense that the
attitude at Demon is: "This is the way we do things and why, the rest of
the world will just have to adjust." E.g.,
 
>Paid to do what ?  Go through everyones mail directories and remove any
>list mail that may bounce ?  [Paid to fix your mail system was the intent.]
 ...
>This would not be relevant since we do not have a limit on the size of
>mail directories. We do not operate a quota system.  [Suggestion for
 "mailbox full" scheme.]
 ...
>We can not do that since we pledge to keep all mail for up to 31 days
>from the date of receipt.
 
Does anyone else see an inconsistency between this last statement and
 
>If subscribers to your list don't read it for weeks
>at a time then why are they subscribed in the first place ?
 
(One could, with only minor restatement, say, "If subscribers to Demon's
services don't use it for weeks at a time then why are they subscribed
in the first place?"  Corollary: Do they really care about their mail?)
 
But to your concrete suggestion:
 
>One idea I have would be not to send warning messages in
>response to mail that has a `Precedence: bulk' header. Would this solve
>problems as far as you can see ?
 
This would work for mailing list software that inserts this header.
Majordomo-operated lists seem to do so (though I believe this can be
configured on a per-list basis, so it is probably not guaranteed).
Listserv-run lists don't seem to, and anyway this would only work if all
list management software adopted the same system (or at least a small
number of different systems) for identifying list-based mail.  I don't
think we are there yet.  Perhaps others on the lstown-l list can add
more insight to this.  Conceptually, this seems to be the direction to
go--as other threads have mentioned in the past, there seem to be
different error acknowledgement levels that are appropriate for
different classes of mail.
 
>I'm not promising anything but it seems to be that this would be a
>workable solution.
 
You don't need to promise anything.  I think the parameters of this
problem are well understood, and I don't intend to thrash it out further
on the lstown-l list.  When and if Demon has a solution in place, I'll
change my policy.  I'm certainly willing to cooperate and make
adjustments _where the software I use gives me the control to do so_
("adjusting" my attitude to the point that I start looking forward to 8-
and 31-day old bounced mail doesn't qualify :-)).  Having pointed out
the problem and protected myself from it, I don't see it as my mission
to change Demon's religion (pun not intended).
 
However, it does seem that Demon has the oddball system (from a mailing
list manager's perspective -- I'm sure they have their own reasons for
it), and it's probably unlikely that too many mailing list managers will
be willing to jump through too many hoops to accommodate it in its
present form.
 
Shahrukh Merchant