GoToLoop

`synchronized` is gonna reasonably slow down access to all of those methods. And concurrency is considered an advanced technique among Processing users. If we decide to apply such techs, it's...

> at java.util.Vector$Itr.next(Vector.java:1137) It's strange this library relies on **_Vector_** container. That class was @Deprecated æons ago due to the fact it was somewhat already `synchronized`! O_o

> ... was to replace these with an _**ArrayList**_ wrapped via **Collections.synchronizedList()**, ... Like I've mentioned, Java deprecated class _**Vector**_ b/c it was unnecessarily `synchronized` everywhere. If that's a desirable...

- Main problem I see in your pull is that you're spreading `synchronized` even to methods w/o loops! - I also see you're applying `synchronized` in order to safeguard _ControllerGroup::controllers_...

> ... is replace the Vectors with (unsynchronized) ArrayLists, and synchronize as necessary in ControllerGroup... That's a much more refined approach. But it can be modified later in some follow-up...

In order for `synchronized` to work, all parts which need it gotta use it. If we `synchronized` our own code, but the library isn't doing as well for the dangerous...

Counter-proposal done. Lotsa refactors there: https://github.com/sojamo/controlp5/pull/73

https://forum.Processing.org/two/discussion/19064/add-border-stroke-to-controlp5-controller-element

https://forum.Processing.org/two/discussion/19065/controlp5-responsiveness-issue

In pure Processing, I'd need to replace callback **keyPressed()** w/ **keyTyped()** in order for "system" variable _key_ to get an accented character. https://Processing.org/reference/keyTyped_.html