RFR: 8370498: Improve how Node detects whether a layout property change requires a new layout pass

[Johan Vos]

On Fri, 24 Oct 2025 15:11:05 GMT, Andy Goryachev <andy.goryachev at oracle.com> wrote:

> would agree that adding a whole battery of functional tests for layout is out of scope and better done as a follow-on.

I'm not sure I agree with this. Personally, I consider the battery of functional tests to be more important than the fix, so without the tests, I would be very conservative.
However, we're still relatively early in the 26 release cycle, and we won't do a 26 LTS, so risks on real-world drama's are limited. We can still revert this PR (if it got merged) and the previous one later in the 26-development phase.

I'll look into a more deterministic test scenario -- the suggestion from Andy with 3 levels make sense (although reality is much more complex, with Platform.runLater() running in between 2 pulses, and *that* is non-deterministic as it depends on processor speed etc.

-------------

PR Comment: https://git.openjdk.org/jfx/pull/1945#issuecomment-3448244897