On Jan 23, 2014, at 10:13 AM, Murray S. Kucherawy <[email protected]> wrote:

> On Thu, Jan 23, 2014 at 9:17 AM, Franck Martin <[email protected]> wrote:
> 
> On Jan 22, 2014, at 8:38 PM, Murray S. Kucherawy <[email protected]> wrote:
> 
> > The feature request asks for a way to whitelist deliveries for which 
> > there's "p=reject" and DMARC fails as long as there's a List-Id: or 
> > List-Post: field on a message.  This is basically a filter bypass feature 
> > that's at the control of the sender, and it seems like a really bad idea as 
> > described.
> >
> > There would need to be more to it than just this.  Can you develop the idea 
> > more?  What should be in the field if it's present?  Should it be tied to 
> > something else?
> >
> 
> It is not at the control of the sender, it is at the control of the receiver.
> 
> The receiver needs to enter the IP of where mailing lists are sending from. 
> It is up to the receiver to decide if it should override the policy. This 
> feature is part of the spec where receivers can override the policy and 
> indicate it in the reports..
> 
> Ah, I misread the request.  I thought the request was just to skip DMARC 
> enforcement if those header fields are present regardless of origin.
> 
> I think John's point holds though, namely that the receiver will have to keep 
> that list pretty current, or find some secure way of doing so automatically.

Yes, this may not be easy in some scenario. The way I found out on how to help, 
is to log each time I reject an email that contains a List-id or list-post 
header. I process these logs once in a while and pick the IPs I want to 
whitelist. Anyhow the point here is to have the capability, as it is part of 
the DMARC spec. How this list is built is left as an exercise ;)

> 
> 
> Because some mailing lists like google-groups send from the same IP as other 
> mail streams, you only want to override the mail that is obviously from a 
> mailing list. The easy indicator I found is the presence of the List-id or 
> list-post header.
> 
> The risk, of course, is that there might be a way to send abusive mail 
> from/via those IP addresses, and simply tack those header fields onto the 
> message.  That bypasses DMARC entirely.  So the receiver enabling this will 
> need to understand how the sender/relay uses those IP addresses and trust 
> that it's unlikely to change.

Sure.

>  
> 
> you could make it more complicated, and tie IPs with the presence of the 
> header and its content  (usually the name of the mailing list), but also it 
> is common that several mailing lists are hosted on the same machine, so I 
> don't see the need to make the criteria of surgical precision.
> 
> There are other possibilities.  For example, List-Id: is generally the list's 
> submission address with the "@" changed to a ".".  You might try saying "If 
> it comes from IP X, the value of this field ought to look like *.Y, where Y 
> is the Organizational Domain Name associated with X".  And maybe the domain 
> name part of that needs also to be DMARC-aligned.  I'm making this up as I 
> go, but you get the idea.

Yes, This is what mailman does, but I think ezmlm (or majordomo?) does not put 
a list-id, only a list-post. What I mean, it is not well standardized, because 
mailing list software is usually old. And I'm on some mailing lists, which 
behave more like an alias than anything else, but then don't break DKIM :P
 

Attachment: signature.asc
Description: Message signed with OpenPGP using GPGMail

_______________________________________________
dmarc-discuss mailing list
[email protected]
http://www.dmarc.org/mailman/listinfo/dmarc-discuss

NOTE: Participating in this list means you agree to the DMARC Note Well terms 
(http://www.dmarc.org/note_well.html)

Reply via email to