Bug #1431 - NewtCanvasAWT broken: Resize not working

Hi,

With version 2.4.0 (still existing in 2.5.0) came a new error/bug. A bug ticket was already raised in march 2023 by someone else (bug number 1431).

Unfortunately the bug is not yet confirmed.

I have the same problem. I can confirm the problem with the example code attached in the bug report and in our real world application as well.

To make it short: The NewtCanvasAWT in a Linux enviroment is not resizeable. If you change the size of the window/canvas the rendering stops.
From my perspective it's a show stopper. Jogl is broken (with NewtCanvasAWT under Linux).
I'm only saying this so clearly because I have the feeling that the relevance of the bug has been underrated so far.

I tried to locate the bug, but I failed. The debug messages gave no hint to me. I think the parenting lost during/after the resize.

There is a kind of workaround, but it's very ugly: Add a resize listener to the frame and in every resize event remove and add the NewtCanvasAWT from/to the component.
You can shorten that by calling only removeNotify and addNotify in the NewtCanvasAWT component. But this isn't a proper solution.

I also noticed that rendering in a multi-monitor environment works again after moving the window to another monitor (but only until the next resize). I think it's because of the the calls of removeNotify/addNotify which are caused by the monitor device change.


I know that time/attention is a valuable commodity in an open source project. Hence I can offer support for fixing the bug. But I would need a hint what/where is the problem.


Details of my environment: Rocky Linux 8, nvidia GPU, JDK 8/11/17

Thank you and foremost thx to crougier
for submitting https://jogamp.org/bugzilla//show_bug.cgi?id=1431

I have raised its priority, might be related to late HiDPI changes.

Yodoga, is the bug description sufficient or do you have another test case?
If so, please send me a PM (private email), get a bugzilla account and add such info.
Thank you.

It will be done .. and I agree, we can't tolerate an issue with anything NEWT.

Sven Gothel wrote

Yodoga, is the bug description sufficient or do you have another test case?

The provided test case in the bug ticket is very good - gets right to the point.

Bug 1431 is resolved