Michael Ennen

This seems like it may be a JavaFX issue? Can you reproduce it with `Platform.runLater`s and not using FxRobot?

Well you are running in headless mode, correct? I just meant check the assertions and see if they still hold without using FxRobot (or any TestFX functionality).

@BiosMono That's more difficult to test without TestFX because you are using `FxRobot.clickOn`, but if you can also try to reproduce without TestFX, that would be helpful. It's hard to...

Okay, then I would have to say that this is a bug in [Monocle](https://github.com/javafxports/openjdk-jfx/tree/develop/modules/javafx.graphics/src/main/java/com/sun/glass/ui/monocle) (the headless JavaFX platform).

> I´m able to reproduce the test case mentioned by @yamidark on linux with Xvfb, where is no Monocle involved. Do you mean where there is no TestFX involved?

I don't understand why you would run with xvfb but without monocle? AFAIU xvfb only needs to be used when running TestFX in headless mode with Monocle.

So `TestFxBase` test fails when using TestFX in headless mode on Linux with Monocle and xvfb? Does it also fail without using JFXPanel? Can you open a new issue?

@yamidark I probably should have looked at this more closely...does the assertion fail if you increase the sleep time ? It *could* just be a timing issue, and 1000 milliseconds...

Probably going to require someone to step up and debug it, I am quite busy for the next few weeks unfortunately.

The `jdeps` tool shows the cases where we use internal APIs: ``` testfx-core-4.0.5-alpha.jar -> C:\Program Files\Java\jdk1.8.0_91\jre\lib\ext\jfxrt.jar org.testfx.service.adapter.impl.GlassRobotAdapter (testfx-core-4.0.5-alpha.jar) -> com.sun.glass.ui.Application JDK internal API (jfxrt.jar) -> com.sun.glass.ui.Pixels JDK internal API (jfxrt.jar)...