|
Neohapsis is currently accepting applications for employment. For more information, please visit our website www.neohapsis.com or email hr@neohapsis.com |
From: Wietse Venema (wietse
porcupine.org)
Date: Sun Jan 27 2008 - 08:29:37 CST
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Tony Earnshaw:
[ Charset ISO-8859-1 unsupported, converting... ]
> Tony Earnshaw skrev, on 22-01-2008 01:36:
>
> > Mark Martinec skrev, on 21-01-2008 17:18:
> >
> >>> I'm running Postfix 2.4.6 (tried 2 different 2.5 snapshots but both
> >>> attempts gave horrible and unacceptable side effects)
>
> >> This is the second time you are saying that 2.5 is broken,
> >> whithout providing any explanation or evidence on what is broken.
> >> I'm running 2.5 snapshots since they exist, and have no problem
> >> interfacing it with dkim-milter or dk-milter, or noticing any other
> >> anomaly.
> >
> > Yep, probably.
>
> I'd better put the record straight, since Postfix 2.5 is now de facto
> standard stable, I have to adopt it whether I like or not and people on
> the dkim-milter list may get the wrong idea.
>
> I've never written that PF "2.5 is broken", I have written that it "gave
> horrible and unacceptable side effects". This is because PF 2.5 has many
> optional parameters that were not available with 2.4. The one in
> particular that made it unacceptable is how prepend_delivered_header is
> implemented. The new options resulted in all messages with Delivered-To
> headers being bounced, whereas that did not necessarily happen with 2.4.
The only Postfix 2.5 change with prepend_delivered_header was that
Delivered-To: loop detection was broken, and now works.
Of you don't want Delivered-To: loop detection in pipe(8), then
don't specify the "D" flag.
Wietse
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]