On 12 Mar 2004 at 13:18, Nick Ragouzis wrote about
"RE: [Exim] Request for comment: cha":
|...
| And if the circumstances call for the message to persist at the
| receiver my argument stands: with certainty the message reception was
| *COMPLETED* once DATA started.
Heh. Well, by that argument, if circumstances call for a message to
persist then "with certainty" message reception was "completed" once
the idea of the message was a gleam in the original sender's eye! Or
keeping it within the SMTP realm on the receiving machine, once the
initial connection was established.
Sorry, getting silly.
| Well, your next question is more to the point, isn't it ...
|
| > Is there any consistency to what other MTAs do?
Yup. Consistency == expected practice, which would be the strongest
argument. But the reason I asked it is that you seem to be positing
automated tools operating on the Received: header timestamp with an
expectation that it represents the start of DATA time. I don't know
what tools you are talking about, but their existence implies there
is consistency among MTAs that this is true.
Given that RFC821, RFC822, and RFC2821 all fail to provide any
semantics for the time stamp more precise than "the date and time the
message was received", I'd find a high degree of consistency quite
surprising.
Cheers!
- Fred
BTW, Philip must get quite a chuckle out of threads like these...
