Mailing List Archive

header protection drafts too early to implement (Re: Protect email experience not Subject:s (hypothesis, draft))
Took a few hours to read through the current version of

Am Freitag 29 Januar 2021 17:52:25 schrieb Bernhard Reiter:
> [3] https://datatracker.ietf.org/doc/draft-ietf-lamps-header-protection/

draft-ietf-lamps-header-protection-03 Last updated 2021-02-22
which also aims at OpenPGP/MIME mails.

To keep you in the loop, my main take-away so far:

It is not ready to be implemented yet, because

a) it rightfully aims to proposed one method and leans towards
wrapped message approach.

b) the usability problems are not addressed, mainly how to display a mixed set
up headers where some are signed-only, not protected or
signed-and-encrypted, but also the complexity arising from this.
Also what should happen if one of the signature do not validate,
displaying this for each header field is something I cannot really
imaging so far.

c) the drawback of server filter and access and indexing problems
with implementations (where email draws a lot of usability from)
are unsolved.

Overall, there is some good technical work in the document. Personally I feel
it focusses on some aspects more than other aspects, missing a bit on the
bigger picture.

Best Regards,
Bernhard

--
www.intevation.de/~bernhard   +49 541 33 508 3-3
Intevation GmbH, Osnabrück, DE; Amtsgericht Osnabrück, HRB 18998
Geschäftsführer Frank Koormann, Bernhard Reiter, Dr. Jan-Oliver Wagner
Re: header protection drafts too early to implement (Re: Protect email experience not Subject:s (hypothesis, draft)) [ In reply to ]
Am Freitag 12 März 2021 18:02:41 schrieb Bernhard Reiter:
> To keep you in the loop, my main take-away so far:
> It is not ready to be implemented yet, because

If it is implemented, to me it makes sense to
a) only implement one method, and this seems to be
to wrap one full message in MIME, because it is most
backwards compatible and proposed in the current draft.

Thunderbird does implement a different outdated approach I believe
(At least I've examined an email from Thunderbird/78.6.0)
and I found Content-Type: multipart/mixed; boundary=".........";
protected-headers="v1")


b) implement reading first and do not activate sending.

This is something that Thunderbird should also fix.

Implemententation for reading makes some sense to try out how the unaddressed
usability problems will fare out.

To send encrypted subjects now means losing information if an injected
approach is used like Thunderbird seems to use.

Best Regards,
Bernhard

--
www.intevation.de/~bernhard   +49 541 33 508 3-3
Intevation GmbH, Osnabrück, DE; Amtsgericht Osnabrück, HRB 18998
Geschäftsführer Frank Koormann, Bernhard Reiter, Dr. Jan-Oliver Wagner