Re: [dmarc-ietf] From: munging, was Ratchets - Disallow PCT 1-99

[Douglas Foster <[e-mail]>]

>
> Ale said:
>
"ARC is not a part of DMARC, despite the acronym being a substring."

Is this really true?  Some time ago, the chairs said that ARC was the
candidate solution to the mailing list problem, and that DMARCbis would not
fly without a mailing list solution.  More recently, we seem to be
proceeding with a plan to leave the based design of DMARCv1 untouched, with
all of its limitations.   So perhaps the politics have changed.

I have tried to fall in love with ARC, without success.   I have not seen
anyone describe an algorithm for parsing an ARC chain to draw a
conclusion, although the effort seems to require an extraordinary amount of
software development and processing effort.  Having the outlines of this
algorithm seems important, if we expect both implementers and evaluators
for ARC.

But even with an algorithm, the goal seems elusive.  Since any liar can
start an ARC chain and insert a set of unverifiable A-R assertions, it
seems that trusting an ARC set requires prior knowledge of whom to trust,
and when that information is available, ARC chain evaluation becomes
redundant.

Doug Foster