From: Noel Jones (njonesmegan.vbhcs.org)
Date: Wed Jun 01 2005 - 12:53:10 CDT

• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

At 11:58 AM 6/1/2005, Ben Goodwin wrote:
>-----BEGIN PGP SIGNED MESSAGE-----
>Hash: SHA1
>
>Instead of having Postfix fill in the gaps when it comes to RFC's and the
>FROM: header, would it be wise to reject the mail after the DATA
>conversation completes if there's no valid FROM: header? Could I use
>header_checks(5) (or are header_checks even done before Postfix adds a
>missing FROM: header)?
>
> -=| Ben

A year or so ago I experimented with rejecting mail with
existing-but-invalid From: headers and found many otherwise valid messages
(mostly mail lists) would be rejected, so was not able to reject based on
badly formed headers. If memory serves me, over half of the messages
detected with bad From: headers were non-spam - far too much collateral
damage to be a useful test.

I strongly suspect the same for missing From: headers, but no way currently
exists to trigger an action based on a missing required
header. Header_checks, which examine each line independently, cannot
detect and are not designed to detect missing headers.

So no, it would probably not be wise to reject mail due to a missing From:
header, and regardless, there is no way to do this without modifying the
source.

--
Noel Jones

• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]