"Jason R. Mastaler" <[EMAIL PROTECTED]> writes:

> "Chris Berry" <[EMAIL PROTECTED]> writes:
> 
> > I'm using this qmail patch
> > http://netdevice.com/qmail/goodrcptto.log.v7.patch which basically
> > requires that user addresses be in the file goodrcptto or
> > moregoodrcptto.cdb before it will allow delivery.
> 
> I still have a hard time believing this patch breaks the "default"
> notion of qmail extension addresses.  That is, mail to jasonrm-foo-123
> and mail to jasonrm-foo-abc will both be handled by
> .qmail-foo-default.  

I downloaded and read through the patch.  It does break extension
addresses.  The exact envelope recipient address must be in the
goodrcptto text file or moregoodrcptto.cdb for it to let the message
through qmail-smtpd.

I was not aware of this patch previous to Chris's post, but it's
really something I need.  I got joe-jobbed pretty thoroughly a couple
of weeks ago and qmail has no native way to prevent that.

I'm going to look at enhancing the patch so that extension addresses
*do* work.  I'm thinking about something similar to the qmail-users
mechanism, which supports wildcards.  That way, you would simply
specify something like

=tim
+tim-

to handle a plain tim@ or any tim-EXT@ address, while still refusing
mail addressed to any non-user of the system.  The patch isn't that
long and there is plenty of code to borrow from qmail-newu and
qmail-local.


Tim
_________________________________________________
tmda-workers mailing list ([EMAIL PROTECTED])
http://tmda.net/lists/listinfo/tmda-workers

Reply via email to