RFR: 8305418: [Linux] Replace obsolete XIM as Input Method Editor [v22]

[Martin Fox]

On Sun, 18 Aug 2024 20:10:09 GMT, Thiago Milczarek Sayao <tsayao at openjdk.org> wrote:

>> modules/javafx.graphics/src/main/java/com/sun/glass/ui/gtk/GtkView.java line 123:
>> 
>>> 121:         if (w != null) {
>>> 122:             pos[0] -= (pos[0] > 0) ? ((w.getX() + getX())) : 0;
>>> 123:             pos[1] -= (pos[1] > 0) ? ((w.getY() + getY())) : 0;
>> 
>> This code will only adjust a position coordinate if it's positive. I can't easily test a dual-display Linux system but I'm pretty sure X or Y can be negative on a secondary display to the left or above the primary display. I think the conditional checks should be removed (but would appreciate a second opinion on this).
>> 
>> By the way, I test on a Retina display at 200% scaling and the IM positioning is working as expected.
>
> I tested on both XWayland and Xorg. They are never negative.
> 
> The input method window did always pop up on the right place.
> 
> ![image](https://github.com/user-attachments/assets/05568b82-3602-48d0-acf1-65aabc8a316b)
> 
> ![image](https://github.com/user-attachments/assets/14d29bc3-9e57-47c3-b573-f08871306c69)

Interesting. I just did a quick test on the Mac and verified that either coordinate can go negative. There must be a difference in the way the platform's model multiple monitors.

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

PR Review Comment: https://git.openjdk.org/jfx/pull/1080#discussion_r1721064868