From: Daniel Stenberg (daniel@haxx.se)
Date: Wed Mar 13 2002 - 04:46:47 CST
On Tue, 12 Mar 2002, Bela Gazdy wrote:
> As I understand, sorting is done by arrival date, which is, most of the
> time, a pretty good method. There is one reference clock, and all others
> are irrelevant; However, in certain situations, such as when the receiving
> machine and/or hypermail is down for a periode of time, or, in case of
> mailing lists, the list server is down, there is no guarantee that a
> 'reply' would not arrive before the original when services restored.
I believe one other major reason why the "Date:" header was rejected is the
fact that users are notoriously bad at setting their local time correct or
even worse their time zone, so if we sort mails based on what all the remote
users' compuers think the time is when they sent the mail, we will get some
very odd results. ;-)
> Would it be too hard to make a config option available to tell hypermail to
> sort using, say, the date obtained from the last "Received: from" header
> line ? I know, this could result in even worse indexes, but at least then
> we could make a more informed decision.
Using the Received: lines sounds like a clever idea, and probably also quite
doable.
I'd say that the most boring part would be to actually parse those lines, as
they tend to come in many different syntaxes and the chance of getting it
right even close to 100% of the times is very slim.
I don't think anyone would mind if someone produced a patch that introduced
this funtionality.
--
Daniel Stenberg - http://daniel.haxx.se - +46-705-44 31 77
ech`echo xiun|tr nu oc|sed 'sx\([sx]\)\([xoi]\)xo un\2\1 is xg'`ol