Hi Murray,
At 14:24 24-01-2008, Murray S. Kucherawy wrote:
>If dkim-filter encounters an Authentication-Results: header it can't
>parse because it doesn't conform to the latest draft, should it:
>
>a) delete such headers
>b) ignore such headers
>c) try at least to determine the "authserv-id" (usually the hostname that
>supposedly added the request) and then decide whether or not to delete it
>
>I'm inclined to implement (a) but be willing to do (b) as an option.  The
>filter currently does (b).

We may have the following headers which were not inserted by mail.example.com:

Authentication-Results: mail.example.com; dkim=pass (1024-bit key) 
[EMAIL PROTECTED]

or

Authentication-Results: mail.example.com; 
[EMAIL PROTECTED]; dkim=pass

For the first one, we might either do (a) or else reject the message 
as we don't consider a message having our authserv-id as valid.

The second case is a malformed header where we can use (b) if we 
assume that the header won't be of a problem to parsers which process 
our Authentication-Results header.  I would not consider this merely 
as a malformed header as it includes our authserv-id.  As such, I 
suggest doing (c) or else rejecting the message.  Note that reject is 
a policy decision here and it's up to the site to decide.

>The risk of doing (b) is that a naive MUA or filter downstream that
>doesn't bother to check the correctness of the header format might
>interpret an older header, which could've been forged but was let through
>dkim-filter because it couldn't be fully parsed for analysis.

Agreed.  People reading the header might also incorrectly assume that 
it was inserted by our MTA.

>I don't particularly like (c) because by the time this is a standard,
>there should only be one form to parse and use by any filtering agent or
>MUA, and thus trying to remember all historical formats or common
>malformations will make the code highly complicated.

I don't like (c) either but I suggest considering as an option as we 
cannot assume that this header won't be used for questionable purposes.

Regards,
-sm 


-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
dkim-milter-discuss mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/dkim-milter-discuss

Reply via email to