[messaging] On Signed-Only Mails

[Moritz Bartl]

On 11/29/2016 10:25 AM, Peter Gutmann wrote:
>> [...] Signed-Only Mails are Useless [...]
> Yup, and it's for exactly the reasons given there that the S/MIME WG decided
> many years ago not to sign messages sent to the list.  Courts, similarly, rule
> on the intent of the signer, not some attached bag of bits (see e.g. Steven
> Mason's excellent "Electronic Signatures in Law").  So while I wouldn't go so
> far as to call them harmful, I'd agree that they're mostly useless, unless
> you're using one to make some special point.  Even then, if it's for legal
> purposes, a court will look at almost everything but the signature when
> deciding on its effect.

This is a dangerous and wrong statement. For one, you are making a case
based purely on previous cases in US courts, which is a very US-centric
view, and dangerous for a discussion that potentially affects all
jurisdictions. Secondly, even within US law, even if you're right in
what you are saying, the absoluteness of your statement makes it wrong.
Courts in most jurisdictions, including the US, will take everything
into account, and if the circumstances and expert witness help support a
case using the signature, the court will not simply ignore it. As an
example, I'm on the board of a US organization which use GPG signatures
explicitly in their bylaws to express a certain meaning. Since we all
aim to resolve conflicts outside of courts (don't we), a valid or
invalid signature will certainly weaken or strengthen sides in the
discussion and help resolve potential conflicts before they end up in court.

I think what makes this discussion difficult is the conflation of two
very separate points. One, which I see as the main point made by Vincent
and Bjarni, are the obvious UX problems with signatures either inline or
as attachments, and there separately for clients with openPGP support
and without.

We should look at the "usefulness" of signatures completely separate
from that point. I would argue that people can be trained to understand
the difference between something that is encrypted, and something that
is signed, even just because they are usually educated to understand a
written signature and the difference to a closed envelope. There's no
doubt that written signatures in society are indeed regularly used in
courts and outside in situation of conflicts as a supporting argument.
Looking at technical communities and their adoption, I know at least the
Debian and Torproject infrastructure relies on signed mails for their
LDAP, as example of it lacking wide adoption not because there is no
good use case.

A lot of the problems OpenPGP carries are because of the fact that it
does not prescribe a specific usage model. Still, this does not mean
that one cannot not define useful models on top of OpenPGP and its
signing packets, and agreeing to certain meanings. Compare again with
written signatures and (oh god) company stamps.

I would be very sad to see signing capabilities go away or become an
expert feature "just" because it is a UX problem for clients who happen
to not support the way signatures are currently attached to a mail. We
should have a discussion about the general usefulness of signatures
first, and separately, hopefully come to an agreement that they have a
lot of potential and good use cases, and then in the next step find a
way to transmit and offer them without confusing end users. Even if
"most" people do not currently get the meaning of digital signatures, I
see it as our obligation to teach them, not "make it simple" and remove
an important feature. Look at Adobe PDF digital signatures, or the usage
of SMIME signatures in enterprise environments, or the various
"upcoming" digital signature laws in western countries.

Similarly, I guess we all agree that there is a (meaningful) difference
between an encrypted and an encrypted and signed mail. These are simply
two separate features supported/provided by the same underlying technology.

Moritz