[ietf-dkim] Straw poll results

[Douglas Otis]

  On 8/9/10 11:41 AM, John R. Levine wrote:
>> You're assuming that how end-users sort list messages is the same
>> as how DKIM verifiers might operate on list messages. Is that a
>> good assumption? Or do you mean something else when you say
>> "sort"?
> I suppose I could go back and specifically ask people how their spam
> filters handle list mail as opposed to how they sort it, but I'd be rather
> surprised if the answers were different.  We're pretty high end users
> here, in my case the filtering and the sorting are intermixed and I doubt
> I'm the only one in that situation.
>
> My point is simple: everyone handles mail from lists using the identity of
> the list, not the identity of the contributor.  Despite three decades of
> experience with mailing lists, arguments to the contrary say "someone
> might" rather than "we do".*
>
> I hope we all agree that it is a waste of our time to design complicated
> mechanisms to solve problems that don't actually exist.
John,

Don't lose sight of ADSP's intended purpose. This addressed a problem 
only significantly affecting a very small percentage of email domains.  
Nevertheless, the experience of financial loss drove forward efforts at 
finding a solution.

The ADSP dkim=all tpa-sig; assertion will not require mailing lists to 
change their practices.   Efforts in resolving the issue rests with the 
targeted domains.

Sorting by the mailing list only works when the recipient subscribes to 
the list, and that the list has indeed issued the message.   The intent 
behind the "complexity" of the TPA-Label scheme is to utilize any 
existing method to ensure the list represents the message source, based 
upon guidance offered by the targeted Author Domain.

No amount of reactive domain reputation foo will provide a solution 
within this problem space.

-Doug