BG> However, despite this bug, in the 3 months I've been using 1.10, it
BG> has *never* falsely detected a dupe here.

RS> Yeah, Paul has had a brain fart on this one. Yes its possible to have
RS> the To: From: Date: and Subject: fields identical, BUT you also have
RS> to have the Time field identical too to have a problem.

BG> Which presumably only means users of PQWK prior to v202 ??

Yeah, exactly.

RS> Even when the message is coming from a QWK system which has no
RS> seconds field, the chances are pretty remote. Possible, coz you can
RS> get that if you proof read the mail, you can edit two in the same
RS> minute, but remote. And the 2 second increment in the early PQWKs is
RS> good protection against that. THEN the only thing which will bite you
RS> is proof reading those two messages and changing them BOTH in the
RS> SAME minute, and they must ALSO be PRECISELY 30 messages apart in the
RS> REP too. Very very unlikely.

BG> As near as damn it to infinitesimal, I'd have said.

Yeah, difficult to put a hard number on it coz its hard to quantify the
risk of the crucial one, two messages edited in the one minute which
match in To From and Subject. Date would normally match if you do daily
mail runs.

RS> Your test isnt actually that conclusive coz by definition the vast
RS> bulk of false dupes would have got dropped before they got to you.

BG> Actually, my test is _very_ accurate, because all the TML stuff I
BG> send to David is bounced on the next call, and dupe detection works
BG> perfectly,

Sure, I just meant that it only applys to AVT and Loc* and PD a very
small subset of echoes world wide.

--- PQWK202