[ietf-dkim] the danger of ADSP, was list vs contributor

[John R. Levine]

> If we all agree that that's a valid characterization of ADSP, I suggest we move to get it downgraded from Proposed Standard to Experimental.  There's certainly a lot of rhetoric suggesting it's an experiment that's in the process of failing, though the experiment is also arguably not complete.

At this point, I think a reasonable message is to tell people that it's 
fine to publish whatever ADSP they think best describes their mail stream.

But don't use it to manage your inbound mail unless you're willing to hand 
control of your mail to people of unknown competence and/or virtue.  Some 
senders will be fine, some will not.  I'm trying to come up with as many 
perverse ADSP scenarios as I can.

R's,
John