From: Mark Martinec (Mark.Martinec+postfixijs.si)
Date: Fri Mar 14 2008 - 19:42:46 CDT

• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

Do not get carried away too far with this business of replacing
a Date header field - it is breaking the semantics of this field
as stated in RFC 2822, and it is solving the wrong problem.

If a time of *reception* matters, look at the date/time of reception
in the Received header field as supplied by your perimeter MX
(and make sure this MX MTA has its clock under control of NTP).

If a time of a mail *composition/mail-submission* matters, your only
choice is to trust whatever data is supplied by the author, unless
a submitter is required to provide a certified timestamp on a
document/message from some trustworthy timestamping service.

  Mark

• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]