[GH-ISSUE #314] A single mouse event "leaks" into chained clients when moving back #164

New issue

Open

opened 2026-05-05 22:12:14 -06:00 by gitea-mirror · 2 comments

gitea-mirror commented

2026-05-05 22:12:14 -06:00

Owner

Originally created by @hoshinolina on GitHub (Jul 27, 2025).
Original GitHub issue: https://github.com/feschber/lan-mouse/issues/314

Let's say I have three chained clients like this:

     +---+
     | C |
     +---+
+---++---+
| A || B |
+---++---+

The mouse is connected to A, which then has a client declaration for B, which then has a client declaration for C.

When the mouse is moved from A to B to C, things work as expected. When the mouse moves back from C to B, that works fine. However, when the mouse moves from B to A, the mouse pointer in C (which at this point is at the bottom of the screen, where it was left after B took control back) moves by an amount that seems to be equal to the mouse event that crossed over from B to A (to the left). In other words, somehow, the mouse events are momentarily directed to C again when crossing back from B to A.

This only happens once. Moving back to B and again to A does not leak any events to C. It's only the first time after leaving C, then leaving B, that an event leaks. To make it happen again, I need to move the mouse all the way to C one more time.

Originally created by @hoshinolina on GitHub (Jul 27, 2025). Original GitHub issue: https://github.com/feschber/lan-mouse/issues/314 Let's say I have three chained clients like this: ``` +---+ | C | +---+ +---++---+ | A || B | +---++---+ ``` The mouse is connected to A, which then has a client declaration for B, which then has a client declaration for C. When the mouse is moved from A to B to C, things work as expected. When the mouse moves back from C to B, that works fine. However, when the mouse moves from B to A, the mouse pointer in C (which at this point is at the bottom of the screen, where it was left after B took control back) moves by an amount that seems to be equal to the mouse event that crossed over from B to A (to the left). In other words, somehow, the mouse events are momentarily directed to C again when crossing back from B to A. This only happens once. Moving back to B and again to A does not leak any events to C. It's only the first time after leaving C, then leaving B, that an event leaks. To make it happen again, I need to move the mouse all the way to C one more time.

gitea-mirror commented

2026-05-05 22:12:15 -06:00

Author

Owner

@feschber commented on GitHub (Oct 29, 2025):

Could you add which backends are being used on the three devices?

@feschber commented on GitHub (Oct 29, 2025): Could you add which backends are being used on the three devices?

gitea-mirror commented

2026-05-05 22:12:17 -06:00

Author

Owner

@hoshinolina commented on GitHub (Nov 2, 2025):

This is running on KDE Plasma Wayland on all three machines, with the default backends, which it reports as:

[2025-11-02T12:03:27Z INFO  input_emulation] using emulation backend: libei
[2025-11-02T12:03:28Z INFO  input_capture] using capture backend: input-capture-portal

@hoshinolina commented on GitHub (Nov 2, 2025): This is running on KDE Plasma Wayland on all three machines, with the default backends, which it reports as: ``` [2025-11-02T12:03:27Z INFO input_emulation] using emulation backend: libei [2025-11-02T12:03:28Z INFO input_capture] using capture backend: input-capture-portal ```