On Apr 25, 2016, at 12:58 PM, Ned Freed
<ned(_dot_)freed(_at_)mrochek(_dot_)com> wrote:
It's also conceivable that a company would offer users the ability to
upload
their public key to the repository.
True, although if it's web mail you're going to need some impressively
complex Javascript to arrange so that the user can read the mail but the
provider can't.
It's a nasty problem even if you're willing to settle for possibly
exposing the message but not the private key.
The nasty bit is trusting someone you don't trust with your private key to
serve you the javascript code your crypto is executing on.
The rest of it is fairly easy. PGP in in-browser javascript is a (multiply)
solved
problem, as is secure local storage for private keys.
And even if you're willing to expose the private key to the webmail system,
there's the interesting issue of how to set things up so your mobile
client shares the same key.
The mobile client generates it's own key pair, then the desktop / web /
whatever / client sends it's key pair via encrypted & signed mail to the
mobile client.
To wander vaguely back towards the thread, that'd be a perfect example
of a single email address having multiple not-quite-equivalent cryptographic
identities; mapping an email address to a single associated key isn't
sufficient.
Cheers,
Steve
_______________________________________________
ietf-smtp mailing list
ietf-smtp(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf-smtp