Selection of algorithms

Another issue where opinions would be very useful is the selection of algorithms.

I have re-implemented most mainstream primitives and hence could continue adding test vectors for them. However, there are a number of algorithms for which I could not find any use cases anymore. Without use cases it makes little sense to spend time on them. The list of primitives that need some use cases are as follows:

* DSA (I haven't seen any use cases of DSA)
* binary elliptic curves (time is probably better spent adding newer curves used for cryptocurrencies)
* Morus (looks like a nice algorithms, I just haven't seen any use cases yet)
* UMAC
* VMAC

There are of course protocols that support cryptographic primitives for a long time, so maybe I'm overlooking something.