Re: [Asrg] Call for contributors to develop `pay per fraudulantmessage` I-D, code

[Amir Herzberg <herzbea at cs.biu.ac.il>]

Barry Shein wrote:

> Although not unspeakable it has that "little" glitch in the idea that
> someone has to decide what is spam and what is not.

I agree that this should be addressed, and in fact tried to do so in the 
design of SICS; some details are in the existing version and some will 
be included in the next version.

Essentially, the party detecting the mis-labeled message (our variant of 
spam) demands compensation from the signer of the label (typically the 
sender or his MTA), either directly (if they have agreement) or via a 
`guaranteed payments CA` (combination of CA and Payment Service Provider).

This is handled efficiently with the approach of `optimistic` protocols, 
i.e. we first ask the sender to confirm if it is mislabeled or not. If 
he admits (e.g. due to zombie), he pays fine x$. If he disagrees, this 
is a `dispute` and the MTA (or the guaranteed payments CA) will evaluate 
this. Yes, this costs extra money... to be added to the fine, i.e. you 
now pay x$+y$ (or if it is determined this was not a justified 
complaint, the complaining party pays y$).

..skip
> (I know "did you read what I WROTE?!", YES I DID, ok?! It still comes

No need to be aggressive/defensive... While I do try to write as clearly 
and completely as I can, and appreciate it when people read it and help 
improve, I am well aware that people don't have enough time to read 
throughly every proposal, so I am not complaining (usually...).

Best, Amir
> down to someone, somewhere having to decide what is spam and making
> someone else agree to that judgement, unless one is a govt-mandated
> monopoly this tends to be a bad business model.)

_______________________________________________
Asrg mailing list
Asrg at ietf.org
https://www1.ietf.org/mailman/listinfo/asrg