Consider further isolation of consensus related code

[davxy]

We already have primitive/consensus and client/consensus folders to keep together consensus related code.

Currently these folders contains code for protocols related to block authoring such as BABE, AURA, PoW and some shared utility code.

I propose to add to these folders everything related to consensus (thus also finality) such as grandpa and beefy crates.

Even though this is not paramount, I can't see any good reason to have these crates floating with the rest of the code.
In general I think is a good practice to logically organize the code.

If there isn't any drawback I can easily take on this trivial chore.

If there isn't any drawback I can easily take on this trivial chore.

(commenting in the role of an outside observer)

I7-refactor Code needs refactoring.

The question is: Can this task ("further isolation of consensus related code") be done by strict refactoring?

(Strict Refactoring: zero functional change)

Because even a strict refactoring can still introduce side-effects (things usually visible only in the machine-code, link long vs. short jumps etc.).

Z0-trivial Writing the issue is of the same difficulty as patching the code.

Refactoring should not be labeled as trivial - because it is not trivial.

(Should have own label, like Z9-The-Art-of-Code)

[davxy]

You are right, refactory is not trivial. But this is mostly about renaming and moving files from one folder to another