[module] Raylib misses mouse clicks

[timo-scaleio]

---

Please, before submitting a new issue verify and check:

• I tested it on latest raylib version from master branch
• I checked there is no similar issue already reported
• I checked the documentation on the wiki
• My code has no errors or misuse of raylib

---

Issue description

While working on a game idea I noticed that mouse clicks don't get registered properly. I wrote a much simplified version of the game loop in Go and in C and observed the same problem.

Basically when you click "swiftly" the code below doesn't always print when the mouse button is pressed. When you hold the button for a bit it does print every time.

Furthermore if you set the frame-rate to 0 (no wait) then it works all the time even with swift presses. I haven't tested if there was actual "rendering load" if it would still work all the time... but in my dumbed down, no-render version it does.

We had a chat with @JeffM2501 in Discord here https://discord.com/channels/426912293134270465/1335478227551912016

The Go library I used packages Raylib-5.5 and I installed Raylib-5.5 through Brew as well for the C version.

Environment

• Chip: Apple M2 Max
• MacOS: 14.6.1 (23G93)

Code Example

Go Version:

package main

import (
	"fmt"
	rl "github.com/gen2brain/raylib-go/raylib"
)

const (
	windowTitle         = "Test"
	defaultScreenWidth  = 1024
	defaultScreenHeight = 720
)

var (
	screenWidth  int32 = defaultScreenWidth
	screenHeight int32 = defaultScreenHeight
)

func watchWindowResize() {
	screenWidth = int32(rl.GetScreenWidth())
	screenHeight = int32(rl.GetScreenHeight())
}

func watchUserInput() {
	if rl.IsMouseButtonPressed(rl.MouseButtonLeft) {
		fmt.Printf("left mouse button pressed\n")
	}

	if rl.IsMouseButtonDown(rl.MouseButtonLeft) {
		fmt.Printf("left mouse button down\n")
	}

	if rl.IsMouseButtonUp(rl.MouseButtonLeft) {
		fmt.Printf("left mouse button up\n")
	}

	if rl.IsMouseButtonReleased(rl.MouseButtonLeft) {
		fmt.Printf("left mouse button released\n")
	}
}

func main() {
	rl.SetConfigFlags(rl.FlagWindowResizable)
	rl.InitWindow(screenWidth, screenHeight, windowTitle)
	rl.SetTargetFPS(60)

	camera := rl.Camera3D{
		Position:   rl.NewVector3(0.0, 0.0, 0.0),
		Target:     rl.NewVector3(0.0, 0.0, 0.0),
		Up:         rl.NewVector3(0.0, 1.0, 0.0),
		Fovy:       45.0,
		Projection: rl.CameraPerspective,
	}

	for !rl.WindowShouldClose() {
		watchWindowResize()
		watchUserInput()

		rl.BeginDrawing()
		rl.ClearBackground(rl.RayWhite)

		rl.BeginMode3D(camera)

		rl.EndMode3D()

		rl.DrawFPS(10, 10)

		rl.EndDrawing()
	}

	rl.CloseWindow()
}

C Version

#include "raylib.h"
#include <stdio.h>

int main(void) {
    const int screenWidth = 1024;
    const int screenHeight = 720;
    InitWindow(screenWidth, screenHeight, "Test");
    SetTargetFPS(60);

    while (!WindowShouldClose()) {

        if (IsMouseButtonPressed(MOUSE_LEFT_BUTTON))
            printf("Left mouse button pressed\n");

        if (IsMouseButtonReleased(MOUSE_LEFT_BUTTON))
            printf("Left mouse button released\n");

        BeginDrawing();

        ClearBackground(RAYWHITE);

        DrawFPS(10, 10);

        EndDrawing();
    }

    CloseWindow();

    return 0;
}

[JeffM2501]

The core issue seems to stem from the polling nature of raylib's input handling in the game loop.
The simplest solution would be add an event queue function for mouse buttons like we have for keys.
GetMouseButtonPressed, just like we have GetKeyPressed. This could provide the list of mouse up and down events that happened during a frame for the users that need that info instead of just the current state.

[veins1]

It seems like it's the same issue as #4591 but with mouse buttons. The solution is the same and as Jeff said

an event queue function for mouse buttons like we have for keys.

is also needed for those who want exact events.

[raysan5]

This potential issue could be addressed polling the inputs manually in a second thread. The provided custom_frame_control example could be used as a reference.

Note that hardly any user has reported having problems with current mouse input system in +11 years and redesigning the mouse input system to address a very specific sounds as overengineering to me and implies more complexity and more maintenance cost.

[timo-scaleio]

Is it really a specific problem though? I think the code is pretty simple and I can reproduce this on any Mac here - haven't ttried Windows. I will try the suggested work-around above... however it takes away from the simplicity of the API. I think the idea Jeff had was pretty simple and inline with current design principles of the library.

…

On Sun, Feb 2, 2025 at 7:35 AM Ray ***@***.***> wrote: This potential issue could be addressed polling the inputs manually in a second thread. The provided custom_frame_control <https://github.com/raysan5/raylib/blob/master/examples/core/core_custom_frame_control.c> example could be used as a reference. Note that hardly any user has reported having problems with current mouse input system in +11 years and redesigning the mouse input system to address a very specific sounds as overengineering to me and implies more complexity and more maintenance cost. — Reply to this email directly, view it on GitHub <#4749 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/A5F46A5XJSDLSHKGGLNQNNL2NYGJHAVCNFSM6AAAAABWKCZNO2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMRZGM3TQNRRGM> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[timo-scaleio]

Hi,

I implemented the solutiuon above but the mouse events are still missed. This doesn't seem to address the same problem I reported.

package main

import (
	"fmt"
	rl "github.com/gen2brain/raylib-go/raylib"
)

const (
	windowTitle         = "Test"
	defaultScreenWidth  = 1280
	defaultScreenHeight = 1024
)

func watchUserInput() {
	if rl.IsMouseButtonPressed(rl.MouseButtonLeft) {
		fmt.Printf("left mouse button pressed\n")
	}

	if rl.IsMouseButtonDown(rl.MouseButtonLeft) {
		fmt.Printf("left mouse button down\n")
	}

	if rl.IsMouseButtonUp(rl.MouseButtonLeft) {
		fmt.Printf("left mouse button up\n")
	}

	if rl.IsMouseButtonReleased(rl.MouseButtonLeft) {
		fmt.Printf("left mouse button released\n")
	}
}

func main() {
	ScreenWidth:=  int32(defaultScreenWidth)
	ScreenHeight:= int32(defaultScreenHeight)

	rl.SetConfigFlags(rl.FlagWindowResizable | rl.FlagBorderlessWindowedMode | rl.FlagWindowUndecorated)
	rl.InitWindow(ScreenWidth, ScreenHeight, windowTitle)
	rl.SetTargetFPS(144)

	var previousTime = time.Now()
	var currentTime = time.Now()

	var updateDrawTime time.Duration = 0.0
	var waitTime float64 = 0.0

	var deltaTime time.Duration = 0
	var timeCounter float64 = 0
	var position rl.Vector3 = rl.Vector3{X: 0, Y: 0, Z: 0}
	var pause bool = false
	var targetFPS int = 60

	Camera:= rl.Camera3D{
		Position:   rl.NewVector3(0, 0, 0),
		Target:     rl.NewVector3(0, 0, 0),
		Up:         rl.NewVector3(0.0, 0, 1.0),
		Fovy:       60.0,
		Projection: rl.CameraPerspective,
	}
	
	for !rl.WindowShouldClose() {
		rl.PollInputEvents()

		watchUserInput()

		if !pause {
			position = rl.Vector3{
				X: position.X + float32(10*deltaTime),
				Y: position.Y + float32(10*deltaTime),
				Z: position.Z + float32(10*deltaTime),
			}
			if position.X >= float32(rl.GetScreenWidth()) {
				position = rl.Vector3{X: 0, Y: 0, Z: 0}
			}
			timeCounter += deltaTime.Seconds()
		}

		rl.BeginDrawing()
		rl.ClearBackground(rl.Black)

		rl.BeginMode3D(Camera)

		rl.EndMode3D()

		rl.DrawFPS(10, 10)

		rl.EndDrawing()

		rl.SwapScreenBuffer()

		currentTime = time.Now()
		updateDrawTime = currentTime.Sub(previousTime)

		if targetFPS > 0 {
			waitTime = float64(1)/float64(targetFPS) - updateDrawTime.Seconds()
			if waitTime > 0.0 {
				rl.WaitTime(waitTime)
				currentTime = time.Now()
				deltaTime = currentTime.Sub(previousTime)
			}
		} else {
			deltaTime = updateDrawTime
		}

		previousTime = currentTime
	}

	rl.CloseWindow()
}

[wizzymore]

On mac this is a bug from applications like Magnet or Rectangle with the function of snapping windows to corners of the screen. They will add a delay to your mouse click before sending it to the window you are pressing on ( for non fullscreen windows ), i think this is where your behaviour is from.

Disabling those features, or disabling them for the specific window type ( your game ) or playing in fullscreen will fix the issue.

[moomerman]

Thanks @wizzymore I use Magnet so I'll take a look at that, but it doesn't really explain why I only have this issue with raylib. I've tried SDL3, GLFW and Sokol and none of those have any problem capturing fast mouse clicks. With raylib @60fps I have to hold down the mouse button for 500ms-1s for a mouse click to register.

[swelham]

I can also confirm that closing down the Magnet app fixed this issue for me.