[Asrg] FeedBack loops

[J.D. Falk]

On 11/11/2008 15:04, "Franck Martin" <franck at avonsys.com> wrote:

> Well a bit of background.
> 
> Is the ARF standard well accepted, I see AOL and a couple use them. I see also
> that hiding the user that reports the spam may not allow for the recipient of
> the report to act correctly.

Very well accepted, yes.  Nearly every complaint feedback loop uses ARF.

> Are there any tools/library available around ARF, to 1) create these reports
> 2) parse/process them

http://wordtothewise.com/resources/arf.html has some.

> Any chance to have these tools build in in major MTAs ? Does it make sense?

I'm not sure what that would achieve.  The actual process is:

1. $A sends a message to $B
2. $B complains to $C (their ISP) about the message (example the "spam"
button in the Yahoo! Mail web UI)
3. $C encapsulates the message as ARF, and sends it to $A (if $A has signed
up to receive that feedback)

Lots of people want to make it more complicated, but it really is that
simple today.

> Could they be an after the fact fbl? let say I want to know if my email got
> delivered,  query an mta, supplying some data like domain/sender IP, and the
> mta sends me any report over last few hours..

No, ARF isn't (as far as I know) used for delivery reports.  It was created
for -- and is primarily used for -- user complaints, as described above.