[ietf-dkim] Why mailing lists should strip DKIM signatures

[John R. Levine]

>>>> sign, and doesn't strip any headers.  So what happened?  Yahoo saw my
>>>> signature and sent the reports to me, which was of course useless
>>>> since I don't run the list.
>
> Not completely useless, right? The message did come from you. If it really
> was spam, sent from your account, you'd be glad Yahoo reported it to you.

Reports of mail sent through freebsd lists are completely useless, since 
there's nothing I can do with them.

Reports of mail that haven't passed through other people's list managers 
are fine, and my scripts deal with them just dandy.

>> The list should certainly sign, but the old signature has to go, since the
>> reputation of a list's mail belongs to the list, not the contributors.
>
> I disagree. I say the reputation is partly the contributor's, partly the
> list's.

Could you produce an actual (as opposed to hypothetical) example of a 
situation where it is useful to filter list mail based on contributor 
signatures that may or may not be present?  I know that every filtering 
rule I've ever made for list mail looked for characteristics of the list.

I do the occasional bozofilter but they look at the From: line.  For 
reasons that should be obvious, once you know the mail is from the list, 
it's be much more reliable to do the bozo filtering on From: rather than 
looking for a second signature.

R's,
John