[p5.js 2.0 Bug Report]: Global function getter is slower than 1.x

[Papershine]

Most appropriate sub-area of p5.js?

• Accessibility
• Color
• Core/Environment/Rendering
• Data
• DOM
• Events
• Image
• IO
• Math
• Typography
• Utilities
• WebGL
• Build process
• Unit testing
• Internationalization
• Friendly errors
• Other (specify if possible)

p5.js version

2.0.3

Web browser and version

Chrome 138.0.7204.169 (Official Build) (arm64)

Operating system

MacOS Sequoia 15.5

Steps to reproduce this

Because of #8013, I ran a couple of performance profiles on the sketch there and noticed something weird - the traces point to get in bindGlobal taking up a lot of time,

This is a screenshot of the profile of the sketch shared by @davepagurek in that thread

1.x	2.x

Notice how the calls to floor and random are significantly slower.

See this test just calling random:

https://editor.p5js.org/Papershine/sketches/Feun-KSzF

On 2.0 it takes around 140ms while on 1.11 it takes around 10ms.

Maybe the call to bind every time the function is called might be causing issues?

p5.js/src/core/main.js

Lines 105 to 110 in 5ef51ca

	get: () => {
	if(typeof this[property] === 'function'){
	return this[property].bind(this);
	}else{
	return this[property];
	}

I'm not sure if this behaviour is intended, but might be worth discussing!

[welcome]

Welcome! 👋 Thanks for opening your first issue here! And to ensure the community is able to respond to your issue, please make sure to fill out the inputs in the issue forms.

For guidance on contributing, check out our contributing guidelines and other resources for contributors.
💬 If you have questions or need support, feel free to join the Processing Foundation Forum or visit our Community page.
📜 Please also review our Code of Conduct to understand our community standards.
Thank You!

[davepagurek]

Could you add p5.disableFriendlyErrors = true to the top of the sketch to see what the performance difference is? Otherwise, we may mostly be benchmarking FES and not the base random() behaviour. Is the difference still significant?

[Papershine]

@davepagurek There's still a 10x difference... Let me see if I can profile the code directly for this sketch

Screen.Recording.2025-08-02.at.12.20.59.PM.mov

[Papershine]

Here's the profiling results, the setup is indeed 10x slower but I'm not sure what the cause exactly is...

2.x:

1.x:

[davepagurek]

Maybe worth testing what the performance is like if instead of Object.defineProperty, we just directly assign to window? That may break some behavior for setters, but just for accessors, it could tell us if avoiding dynamic getters entirely matches the performance of 1.x. Another thing we could look into is what global binding looked like in 1.x for comparison -- did we switch up how it's done?

[quinton-ashley]

Can confirm that the performance of p5.js v2 is significantly worse than p5.js v1.

I made the mistake of trying to use getters for this purpose in q5.js v1 and it was surprisingly slow indeed.

It's much faster to directly assign to the window and then for props like frameCount uses JS proxy instead.

https://github.com/q5js/q5.js/blob/3bca0888e37476ac90fda685236061846ecfc8aa/src/q5-core.js#L35