The question isn't if it would work or not; the question is rather:
would /anyone/ aware about crypto sensitivity think that it's a good
idea to mindlessly let a converter be responsible for mishaps rather
than humans? It wouldn't be pretty code, that's for sure. Also the
problem it would solve is already solved -- we already have asymmetric
cryptography in the browser, after all.
As for anything invalidating OpenPGP.js: that will inevitebly
<http://www.w3.org/2012/webcrypto/WebCryptoAPI/> happen
<https://github.com/polycrypt/polycrypt> either way. All hail evolution! :-)
/N
On 2013-01-15 08:21, Sean Colyer wrote:
I came across https://github.com/kripken/emscripten/wiki somewhat
recently. It compiles LLVM code into JS. I would be curious how well
it would work to compile gpg into JS?
I don't think it would invalidate what we do here, because I think
OpenPGP.js has done a lot already and provides a good framework for
continued development.
But I think it might be able to provide some level of insight. I'm
hoping to make some progress on Mymail-crypt for Gmail and OpenPGP.js
soon, but if someone wanted to give this a shot that would be pretty
awesome.
Sean
_______________________________________________
http://openpgpjs.org
_______________________________________________
http://openpgpjs.org
