On Tue, 24 Feb 2009, Steve Kemp wrote:
> On Tue Feb 24, 2009 at 12:22:13 +0000, Alain Williams wrote:
>
> > 2009-02-24 11:49:52 1LbvnM-0002Zg-Su H=lon-gs3dmrelay.mistral.net
> > [217.154.246.188] F=<jonb@???> rejected after DATA: RFC2822
> > says that all mail SHOULD have a Message-ID header.
>
> > Just a request for comments, should I be doing this sort of thing ?
>
> If you want to be RFC-compliant you shouldn't.
>
> Message-ID is not a mandatory header, so you cannot rely upon it
> being present. For a while I was doing something similar and found
> too many false rejections to use it, even though most rejections were
> spam not all of them were.
If you want to add a Message-Id for your own trackng purposes, I used to
have the following clause enabled.
I don't any longer, removed because (I think) then SpamAssassin can add a
bit of prejudice to messages without it -- although it seems the SA score
is 0 now anyway:
## Make sure a message has a valid Message-Id
## Similar to the above; this fixes up the problem if that is desirable
warn condition = ${if !def:h_Message-ID: {1}}
message = Message-ID: <E$message_exim_id@$primary_hostname>
> On the other hand I've had good success rejecting messages with
> no Date header - as that MUST be present...
I've often wondered whether to bother with that. SA assigns a couple of
points for it, but there is certainly more justification for rejecting
outright on that basis. Nearly 500 hits on SA's MISSING_DATE rule today
...
Jethro.
--
. . . . . . . . . . . . . . . . . . . . . . . . .
Jethro R Binks
Computing Officer, IT Services, University Of Strathclyde, Glasgow, UK
