PE>> Can you guarantee that in the last 3 years there was no chance of you
PE>> generating two MSGIDs the same?  MSGID is useless IMO, it is far too
PE>> hard to conform to the spec, and since no-one bothers to conform to
PE>> it, it means that mailprocessors can delete messages at random.

MS> Well although most MSGIDs seem to incorporate a 32-bit integer, I
MS> really don't think anyone is going to care if an ID generated is the
MS> same as one generated 3 years ago.  If someone can prove that, then
MS> they're probably

PE> Well I take standards a bit more seriously than that.  If you're not
PE> going to bother conforming to the standard, you should probably
PE> create a different standard, which says "it doesn't matter if you
PE> accidentally produce duplicate MSGIDs".

In other words Paul is a fundamentalist religious zealot as far as
standards are concerned. He is even into burning heretics at the stake
too |-)

--- PQWK202