[Cryptography] Email encryption for the wider public

[Henry Augustus Chamberlain]

> People are human, and screw up from time to time.  Your scheme would
> work OK if people were perfect, but they're not.  A scheme with no
> provision for account recovery or addresses people can remember is
> unlikely to be usable outside the tiny niches where people already use
> S/MIME or PGP.

I think that's a property of PGP and public key encryption in general:
the private key is what identifies you, and if you lose it then that's
it. I can't solve that problem, but I thought that if you remove
passwords (see my original proposal) and only use private keys, then
users are more likely to appreciate the importance of their private
key, and are more likely to back it up (which can be as simple as
printing out a QR code). PGP is currently very niche, and my proposal
doesn't make encrypted email as easy as ordinary email, I think it
might make it much more intuitive, and so appeal to a wider audience.

> The DANE S/MIME approach is pretty simple (after you wave your hands
> and assume that system managers will implement DNSSEC anyway.)  The
> key for each address is stored in a SMIMEA record in the DNS, so if you
> want the signing or verification key for an address, you just do a DNS
> lookup.  You can revoke a key by removing it from the DNS, or roll
> your keys by adding a record for the new key and later removing the
> record for the old key.  The guts are S/MIME minus the PKI, which
> works pretty well now.

I'm probably missing something here, but doesn't that basically just
mean using the DNS system as the PKI? Not that it's a bad move - I
imagine that the DNS system is probably more trustworthy than some
random key server! In any case, from what I understand, normal S/MIME
uses a centralised trusted system for PKI - that's certainly better
than unencrypted emails, and I hope it takes off, either as in its
current form or with the modifications you mentioned. I think PGP
would then only be needed in cases where you want to avoid the central
authority.

I think the scope and intentions of my proposal might still be
unclear: I'm not trying to solve all the problems of the normal email
system; instead, I'm trying to solve a few of the problems that PGP
has.  I really think that my proposal has all of the advantages of
PGP, while removing some (but not all!) of the disadvantages that it
has compared to normal emails.

Regards,

Henry