[Maintenance]: Reduce use of futures in library code

[kwvanderlinde]

Describe the problem

The majority of the Library API returns futures, even for basic information about the library. This makes it cumbersome to consume libraries, as the caller need to unwrap the future by "blocking" and handling various failure modes that don't actually exist. Meanwhile, for many of these methods, all implementations simply return CompletableFuture.completedFuture(...) as this is plain data associated with the library.

The improvement you'd like to see

Any static metadata (name, namespace, version, etc.) is directly accessible without requiring futures. I.e., instead of:

public interface Library {
  CompletableFuture<String> getVersion();

  CompletableFuture<String> getNamespace();

  CompletableFuture<String> getName();

  // etc.
}

we would have:

public interface Library {
  String getVersion();

  String getNamespace();

  String getName();

  // etc.
}

In general, if all of a method's implementations return CompletableFuture.completedFuture(...), it probably doesn't need to return a future at all. Other methods, such as Library#read(URL) can remain future-returning.

Expected Benefits

It will be easier to write, read, and maintain code that consumes libraries. It also won't appear as though we have UI-blocking code when that's not the reality.

Additional Context

No response

[cwisniew]

Remember, the Library is an interface, and any implementation, including future ones, are required to be thread-safe and usable from any thread, not just MTScript Macros (GUI thread). While namespace, name, and version are probably I would err on the side of moving too little rather than too much