Mouse wheel event in passthrought vs direct usb

[paname]

Code: Select all

 
#pragma METAINFO("base", 1, 0, "paname")
 
init {
mousemapping();
keymapping();
 
}
 
main {
mouse_passthru();
key_passthru();
}
 

using this simple script you can check that the scroll line number differ by a factor of 4 compared to directly attaching the same mouse to the pc. this makes some function break in games like weapon swap getting 4 events instead of one.

you can change the settings in windows to move only a single line per scroll to better exhibit the issue.

try with an excel file with the row numbering you will notice that via t2 it scroll 4 times faster than via direct USB.

OS: windows 10

Device physical setup:
Input A = g903 mouse wireless dongle but same with wired g502
Input B = none
Output = PC multihid
Prog = PC

I also put it on github : https://github.com/J2Kbr/TitanTwo/issues/121

[J2Kbr]

In theory this shouldn't be happening, as when mouse_passthru() is called, it gets the raw values from the connected mouse, without any modification (not considering modifications later performed via script).

Thank you for adding to GitHub, I will be testing and debugging this soon.

[paname]

In theory this shouldn't be happening, as when mouse_passthru() is called, it gets the raw values from the connected mouse, without any modification (not considering modifications later performed via script).

Thank you for adding to GitHub, I will be testing and debugging this soon.

thanks. Waiting for your feedback on this, I have a workaround sort of around this behaviour but it is not clean.

[J2Kbr]

Please let me know about your workaround (can be via PM if your prefer).

[paname]

Please let me know about your workaround (can be via PM if your prefer).

my workaround is just to catch the wheel event and use a combo like this:

Code: Select all

 
 
init {
    const uint8 mmap[] = { 
        MWHEEL_FORWARD,     PS4_R3, 
        MWHEEL_BACKWARD,    PS4_TRIANGLE,
    };
    mousemapping(mmap);
 
}
main {
             mouse_passthru();
             key_passthru();
 
        // wheel fix
 
            if (is_active(PS4_TRIANGLE)) {
                mouse_set(MOUSE_WHEEL,0);
                combo_run(WheelDownFix);
                }
            if (is_active(PS4_R3)) {
                mouse_set(MOUSE_WHEEL,0);
                combo_run(WheelUpFix);
            }
}
 
combo WheelDownFix {
    mouse_set(MOUSE_WHEEL , -1);
    wait(30);
    mouse_set(MOUSE_WHEEL , 0);
}
combo WheelUpFix {
    mouse_set(MOUSE_WHEEL , 1);
    wait(30);
    mouse_set(MOUSE_WHEEL , 0);
}
 
 

the wait time in the combo can even be wait(0) it also fix the issue.

if I don't use this "fix" for every wheel move it sends 4 moves instead of 1 when the mouse is directly connected to PC usb.
it still send longer than what raw should be but at least my game isn't confused by multiple wheel events.

[J2Kbr]

Thanks. With important tasks done in the update released today, now I will be able focus on the GitHub issues, including this.

[J2Kbr]

Found the issue, fixed on firmware 0.98B-36. Thanks for reporting.

[paname]

Found the issue, fixed on firmware 0.98B-36. Thanks for reporting.