On Mar 15, 2014, at 1:40 PM, Nicholas Cole
<nicholas(_dot_)cole(_at_)gmail(_dot_)com> wrote:
And thinking about it even further, it compounds a problem that
someone (was it you, Jon?) has written about in the past. Even though
we all know that key UIDs can be signed by complete strangers, users
are *often* disconcerted by this fact (which is why there is a
no-modifier flag, even if keyservers have never respected it and even
if it would make the use of OpenPGP even more complicated). Still, a
naive user of an OpenPGP program may draw incorrect inferences about
social relationships from UID signatures. Imagine the outcry of users
if they discovered that documents were in the wild that 'might' have
been signed by them...
Yes, I'm probably the person. I created the no-modify and other properties of
2440 and 4880 precisely because it was something that I saw as a barrier to
OpenPGP adoption and a personal peeve of mine. (Also, at an IETF meeting that
happened to be on April First, I did an April Fools OpenPGP presentation where
I presented the anti-identity signature, whereby if enough people over a
threshold signed an anti-identity signature, you'd lose your identity and they
could give it to someone else. That was also an expression of my peeve in this
space.)
Jon
_______________________________________________
openpgp mailing list
openpgp(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/openpgp