BV>> I'm sure that silly mail-tossers and brain-dead dupe detection
BV>> routines are probably the biggest cause for problems,

RS> Well, I think we have adequately eliminated dupe checkers from the
RS> general problem. Thats not to say it cant be the problem in isolated
RS> bits of the network. But it has been shown to not be a general
RS> problem.

PE> How did you eliminate dupe checkers?

By sending messages at the one time with everything the same except the
body of the message. When they both get thru, cant be an aggressive dupe
checker which is too stupid to say include the body of the message in
the check. There was one occasion where that was seen to happen, but it
was on messages directly being lost on a particular BBS they were
uploaded to, not the mail system in general.

And the one of Russells that got lost from a batch of 6 or so was the
only one unique in that batch as far as the subject was concerned, one
about compressing a gas. So it cant have vanished due to a dupe checker.

BV>> although I was just thinking that the FREQ thing could also explain
BV>> part of the problem.

RS> I cant see that its even slightly possible. The only remaining
RS> potential problem is the one Paul saw, you FREQ off the system you
RS> normally get your mail feed from. Then you certainly will get an
RS> unscheduled mail bundle. But you also have to ignore it too once its
RS> received. Presumably many systems would gracefully handle the mail
RS> bundle auto even if it spontaneously got sent with a FREQ anyway.

PE> If I FREQ from Dave, I normally always get mail with it, it's a
PE> natural thing to happen.  And it gets processed, not ignored.

Sure, no argument there. Brenton was talking about what happened if it
wasnt processed in that case. I can see that it might just be possible
with some mailers for that to happen, particularly if you dont have it
setup right.

--- PQWK200