I’m enjoying this exchange tremendously, but Phill just said something that
overcame my reticence to pipe up.
On Mar 17, 2015, at 09:44, Phillip Hallam-Baker <hallam(_at_)gmail(_dot_)com>
wrote:
The body of any modern email is going to be UTF8 after all.
There’s a fair bit of email still being exchanged that isn't UTF8, and I expect
this’ll continue for as long as we’re still using SMTP. Everyone really should
read <http://en.wikipedia.org/wiki/Mojibake> in the context of making these
types of assumptions. I still wish the charset header had been added to OpenPGP
sooner, and made to have teeth.
When we started work on End-To-End, it’d’ve been easier in many ways to start
from scratch, invent our own spiffy new protocol, and evade the tyranny of the
installed base entirely. But interop matters. As do maturity and
understood-ness of protocols. We made some implementation-specific choices (eg
only ever generating ECC keys), but we very concertedly set out not to break
anything. David’s declaration of intent to deprecate is worded a bit more
strongly than I would’ve done, but I certainly don't disagree with the idea of
considered algorithmic tidying in order to encourage modern OpenPGP
implementations to generate contemporarily appropriate keys and messages.
When it comes to decryption, I’m (unsurprisingly) in vigorous agreement with
Derek that this is a software problem, not a human problem. The question is
whether we need a more usable universally backward compatible OpenPGP decryptor
for our older data.
s.
_______________________________________________
openpgp mailing list
openpgp(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/openpgp