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

BREAKING CHANGE: This is a significant licensing change. Please review.

Without this change, this plugin pulls in dependencies for Bevy features I don't need when used as a plugin (E.g. the audio feature is enabled even though neither I nor this plugin require it.) All features are re-enabled in `[dev-dependencies]` so examples are unconstrained.

* 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