[feedback-report] ARF working group interest?

[John R. Levine]

>> This is a good point.  FBL data (other than AOL) tends to be unredacted 
>> because you know who you're sending it to.
>
> The last thing one would do is to confirm users' addresses, or betray 
> honeypots. Will guidance for munging be part of the ARF spec?

Sorry, but that's just wrong.  Several large ISPs send me fully unredacted 
copies of messages in FBL reports.

> I see no contradiction here: either users push a button to generate ARF 
> reports for their own ESPs (updated clients), or move spam messages to ad-hoc 
> IMAP folders (problematic clients) so that an ARF header can be generated 
> directly from the file (any added delay may help anonymization.)

I presume you mean ISP, not ESP.  In my experience abuse reports from end 
users are almost always useless.  Senders will often trust a report from 
the ISP but not from the end user since they'll have an FBL agreement with 
the ISP.

> Should ESPs verify that the stuff is actually spam or just forward it?

Wow, talk about a research topic.

R's,
John