At 9:22 PM -0400 10/20/2001, Ronald F. Guilmette is rumored to have typed:
> So I gather then that for you at least, it might be considered useful for
> a mail differentiation tool to be able to differentiate auto-responses
> from non-deliverable bounces, yes?
Not at all, thanks, I never said that. I wouldn't be interested in such
a thing at all, since I have a 100% accuracy rate differenciating manually,
something impossible for any computer program (not to doubt your numbers,
but I find it tough to believe you could catch the goofy vacation
autoresponders I've seen without a human intelligence behind the
decision-making).
Charlie (who is on vacation, and because he is using a dialup
through UUNET, cannot email you directly becuse of
your server blocks)
