David Dahl

Why does this PR affect 1721 files? I would also like this issue resolved but this PR seems broken to me

nulltxt handles this better: you can create as many keypairs per domain as you would like, each is accessed via a unique ID.

I/O is hard! Thanks!

I need to do a lot of documentation. the basic premise here is not to be a product but to offer an implementation / prototype for browser vendors to consider...

I would be happy to help - the focus I have right now is on Firefox as all of the code i am writing depends on NSS - which is...

I am really not too concerned about supporting all browsers. What I'd rather do is provide a robust prototype in firefox that is a great model for other browsers to...

> Even if only used for salting the password hash as opposed to generating the underlying pair... I just worry about these things: https://github.com/amark/gun/blob/8e4128b474987fe59e2d73e38c56e30b8a4aa291/sea/auth.js#L93 The comment even says `pseudo-random` and...

I pulled in node-forge directly to use the `pki` and `pem` modules. Will need to keep this open to figure out ASCII armor of *signatures* next

My latest discovery is that - even using node-forge when one converts the ascii armored (PEM) public key back to an `RsaPublicKey`, the verification fails as the key length is...

Ok, we now have merged in #52 which handles converting to PEM encoded public key and back to RSA key that can be used for `verification` - [x] JWK ->...