I am somewhat hesitant to rely upon VERP's because I have seen several
problems in how some email programs handle the = sign in the VERP. We all
know that the = sign is a valid character for the local part of an email
address according to the RFC, but there are many mail programs (definitely
some clients, but how MTA's handle VERP's is unknown to me) which incorrect
parse the = sign. The incorrect parsing results in something that looks
like:
list-action-username=userhost@listhost
being interpreted as:
userhost@listhost
Has anyone else found problems with VERP's? I know that I have experienced
several problems with mailto: and reply links when using ezmlm. Some
clients that I know have problems are Yahoo, Excite, etc.
In the short term, I have modified my bounce processor to parse DSN's and
qmail bounce reports.
Thoughts?
Thanks.
James
-----Original Message-----
From: Sean Reifschneider [mailto:[EMAIL PROTECTED]]
Sent: Tuesday, December 12, 2000 1:08 AM
To: James Morgenstein
Cc: qmail list
Subject: Re: Qmail and RFC1894 - Delivery Status Notifications
On Mon, Dec 11, 2000 at 10:04:14PM -0500, James Morgenstein wrote:
>This appears to be used by most of the public mail servers that I have
>tested against, but when a mail bounces out of one of my local qmail
The problem with DSN is that *EVERY* machine that the message passes through
must support DSN, or it fails. QMail doesn't support DSN (unless there's
a patch, you have looked at www.qmail.org, right?). Check out
VERPs -- Variable Envelope Return Paths. Searching google
should provide some good hits.
Sean
--
I never thought I would live in a country which had a
self-declared president.
Sean Reifschneider, Inimitably Superfluous <[EMAIL PROTECTED]>
tummy.com - Linux Consulting since 1995. Qmail, KRUD, Firewalls, Python
