This will make a lot of simpler actions and scorers much less boilerplatey

This allows for easier continuous actions, and I think is more intuitive

Fixes: https://github.com/zkat/big-brain/issues/29

Fixes: #26

* add test for steps and fixes
* define system ordering

* Added EvaluatingScorer

* fixed doc issue

Co-authored-by: doomy <2640792-_doomy@users.noreply.gitlab.com>

Fixes: https://github.com/zkat/big-brain/issues/22

BREAKING CHANGE: All composite scorers now use push, instead of a weird push/when mixture. Additionally, they were previously (erroneously) returning &mut Self from push/when. This is now just Self, which means they will actually work now.

Co-authored-by: doomy <2640792-_doomy@users.noreply.gitlab.com>

Co-authored-by: doomy <2640792-_doomy@users.noreply.gitlab.com>

BREAKING CHANGE: Please adjust your scorers to only work with 0.0 to 1.0 values

This was, in fact, completely broken for switching actions.

This is... ridiculous, but it happend to be required by the
Bevy Parenting system. Without this, the Thinker will be randomly
removed from the hierarchy :(

Fixes: https://github.com/zkat/big-brain/issues/10

Fixes: https://github.com/zkat/big-brain/issues/13

Fixes: https://github.com/zkat/big-brain/issues/12

I *think* this fixes it, but I need to test it more
thoroughly and make sure everything is actually ok

Fixes: #7