lmctv

@reaperhulk @pyca/pynacl-core I'm thinking a possible way forward could be to postpone #523 and release 1.4. In this case, I'd just like to add some notes about the pending deprecation...

@dmerejkowsky would you mind giving some more details about the raw keypairs you are dealing with, and in particular the environment in which they have been generated in the first...

@dmerejkowsky Thank you very much for following up! The reason for asking whether you were in possession of the verification key or not was to understand if an API requiring...

To my knowledge, the hsalsa and its sibling hchacha key derivation primitives are only used as building blocks for the xsalsa and xchacha constructions. While I don't think anyone would...

I think this is mostly a documentation problem, and I'd reccomend *against* any kind of direct secret key [de-]serialization. Instead, we should point users to generate and serialize a random...

libsodium expectation is the caller will send the public part of a correctly generated key-pair, I think the behaviour is correct both on libsodium's side and on our one. I'm...

Since all of the exceptions PyNaCl raises are subclasses of nacl.exceptions.CryptoError (https://pynacl.readthedocs.io/en/latest/exceptions/), downstream users can (and IMHO should) wrap any call into nacl code inside a try:..except nacl.exceptions.CryptoError: ... control...

`six` and `cffi` are direct requirements of `pynacl`; without some information on both the build and the deploy target environments and possibly the exception backtrace we are unable to understand...

@kubinski could you please retry your deployment, to confirm #485 really solves this issue?

First, I've just edited your comment, inserting a couple of ` ``` ` markers to get readable output from GH's markdown renderer.