Nathan Sweet

John's point makes some sense, that marking it as "applied" (especially when it really has not been) is a roundabout way of preventing it from being replaced. Still, the reason...

> What about events still hanging in the queue? AnimationStateListener events are only queued within AnimationState methods. It should always be drained before returning. > Also, what about manual changes...

The `timeToWait` setting is not good for asynchronous behavior. If I want to call a method and continue on immediately, I have to set `timeToWait` to 1 and catch the...

Seems like `getScaledInstance` with `SCALE_AREA_AVERAGING` is the only sure way to get better results than bilinear.

Issues are too large, it's terribly distracting. We can't use the GitHub integration at all because of this.

Months go by and this continues to be terrible.

It's pretty obvious by now that no one cares.

Can you provide example code that shows this problem?

Combining multiple fixes in the same PR makes review difficult. Please explain the fixes.

Splitting it into multiple PRs would help a lot.