clock.getDelta() sometimes returns 0 which leads to division by 0 issues

[pmi-gbs]

Describe the bug

I've noticed that in the EnvironmentControls, which broke the camera transform (NaN and the likes) and other values of it.

in:

update( deltaTime = Math.min( this.clock.getDelta(), 64 / 1000 ) )

It should probably defused right after this line (and other lines where a deltaTime is used) with something like:

deltaTime = Math.max( deltaTime, 0.000001 );

To Reproduce

Steps to reproduce the behavior:

Not sure, to be honest, took some time to find out, that getDelta() is allowed to return 0, it randomly occured, while moving around with the controls. It could be that update() was executed consecutively, but even then I don't think it should break the camera.

[gkjohnson]

Thanks - are you passing "deltaTime" in yourself? Or relying on the built in clock? Either way after testing by directly passing 0 in in a second update call it looks like this is causing an issue. I think an easy solution here is to just return early from the update function if is deltaTime <= 0 in EnvironmentControls and GlobeControls. Would you like to make a PR?

[gkjohnson]

Fixed in #743. Without any more information about how this is happening in practice (update must be getting called back to back) checking for 0 seems like the most reasonable solution.

[pmi-gbs]

Yes it did happen with the default clock and my own clock as well. I wonder why though, as even consecutive calls of clock.getDelta() should intuitively return a value > 0, but I guess the precision of performance.now() is kinda bad (and because of all the security stuff got even worse up to > 100 us)

The merged fix is even better thanks. I checked the rest of the code-base and indeed these were the only concerning clock.getDelta() calls (as otherwise only delta * ... is done).