Joshua Nichols [EMAIL PROTECTED] wrote:
Charles Cazabon [EMAIL PROTECTED] wrote:
No need for a special header; there are options to qmail-inject
to make qmail
use per-recipient VERP.
Handling the bounces is up to you; in essence, you create a
.qmail-something-default file and pipe the messages to a script of your
choosing. In your case, you'd just log the $DEFAULT portion to a file.
What I'd like to do though, is take advantage of the ~20 day testing that
ezmlm provides. I don't want to end up unsubscribing people because of an
Out of office message or a temporary errors like mailbox full unless
they've been occurring for a few weeks.
Your previous message seemed to imply you wanted to do this manually, based on
a record of what recipients bounced. That's why I suggested just logging
$DEFAULT.
If I understand the documentation correctly, VERP will just add the
recipient address (or certain other info) to the envelope, and I don't see
how that helps me. If I set up a .qmail-something file, and then piped all
the messages through a script, wouldn't that catch ALL messages (misdirected
unsubscribes, out of office, mailbox full, delivery delays, etc.)?
Not all -- it only appears in the envelope sender. People are extremely
unlikely to send to that address with unsubscription requests (although that
would actually _help_ you), etc. Autoresponders are of course a problem, but
if your message appears to be a mailing list message (Precedence: bulk, etc
headers), any well-written autoresponder will not respond to it.
Note that 50% of the autoresponders out there are _not_ well-written.
Or, are you saying that duplicating this is just a matter of using VERP and
writing my own ezmlm-warn and ezmlm-return? Heh, just...
If you want to use ezmlm, you can edit the appropriate script file to only
remove the automatically unsubscribe after the probe bounces step, and
replace that with a step that logs the username or emails you, or electrifies
your chair. Your choice.
Charles
--
---
Charles Cazabon[EMAIL PROTECTED]
GPL'ed software available at: http://www.qcc.sk.ca/~charlesc/software/
Any opinions expressed are just that -- my opinions.
---