glv2

> Frankly, I have no idea how to go about it. Splitting it into `ironclad.digest` and `ironclad.cypher` would be a good start, but more than half of the digests are...

> Here's a lightly tested first try: [luismbo@879ab77](https://github.com/luismbo/ironclad/commit/879ab77dddf5701e881791141ba3d03d0c57798e) I took a look at your patch, here are a few notes: ```keystream``` is not a cipher, it's a function to jump...

Yes, and ```ironclad/digests``` would load all the digests. These would in fact be almost empty system definitions with a long list of dependencies.

@luismbo I created a [split-systems branch](https://github.com/sharplispers/ironclad/tree/split-systems) with some fixes on top of your patch. When loading the global ```ironclad``` system, it looks like things work fine and the tests pass....

I guess the best solution will be for each subsystem to have its tests, and the test target for the global ```ironclad``` system will run the tests of all the...

Subsystems are available as experimental feature in Ironclad 0.56.

With the subsystems in Ironclad 0.56, you can do ```(asdf:load-system "ironclad/digest/sha256")``` and it will only load what is necessary for SHA256. I'm not really a fan of the mandatory "one...

Currently Ironclad has PKCS1-v2.1, but not PKCS1-v1.5. However maybe a variant of the ```add-openssl-padding``` example in issue #11 might be enough for your use case.

As Ironclad usually includes basic cryptographic primitives and not complete protocols, maybe we could add functions for the raw PKCS1 v1.5 encoding, encryption, decryption, signature and verification as it is...

Currently keeping track of time is not implemented. However it is possible to implement the DHT protocol without it. I did it for the DHT11 sensor in https://github.com/glv2/microscheme/tree/dht (tested with...