Re: [lamps] Robert Wilton's Yes on draft-ietf-lamps-e2e-mail-guidance-15: (with COMMENT)

[Daniel Kahn Gillmor <[e-mail]>]

Hi Robert--

Thanks for your review of draft-ietf-lamps-e2e-mail-guidance!

I've addressed most of your nits in
https://gitlab.com/dkg/e2e-mail-guidance/-/merge_requests/13, which we
should be able to merge soon.

A brief reply below about something i offered no concrete change for:

On Mon 2024-03-04 02:47:14 -0800, Robert Wilton via Datatracker wrote:
> (2) p 26, sec 6.4.  Signature failures
>
>    A conformant MUA MUST NOT render a message with a failed signature as
>    more dangerous or more dubious than a comparable message without any
>    signature at all.  In both cases, the Cryptographic Summary should be
>    Unprotected.
>
> Does it still make sense to flag the failed signature at all, e.g., so
> potentially the receiver can warn the sender that their signature is failing?

I love the idea of enabling useful feedback for this kind of thing.

But the question about what to do for a failed signature seems wrapped
up tightly with what to do with a *missing* signature.  Surely if an
attacker wants to break a signature without alerting the sender of the
breakage, they can just strip it, to avoid the warning.

As such, a streamlined means for providing feedback about failed or
missing signatures probably needs to be designed around a mechanism
related to whether to *expect* a signature in the first place.

You can see a sketch of this kind of work at
https://datatracker.ietf.org/doc/draft-dkg-lamps-expect-signed-mail/ ,
which Aron Wussler and i presented at SECDISPATCH in IETF 118.  But
there are some thorny questions in there, and it's not at a stage where
it can be incorporated clearly in the e2e-mail-guidance document, though
the "Future Work" section does explicitly call for more work on
"Expectations of Cryptographic Protection".  If you (or anyone) are
interested in this work, it would be great to get more feedback on it.

Regards,

        --dkg