Live data from Hacker News

JavaScript Physics Engine

code.google.com

11–20 of 24 posts

Re: JavaScript Physics Engine

#11

I open sourced my "toy" version of JavaScript physics engine just few days ago too. It's not really for serious work, but may be (hopefully) simple and good enough for beginning programmers to learn one way of starting such a project. https://github.com/fuzzthink/SPE.js

It looks good. Have you tested it with SVG? I've been using https://github.com/hrj/box2d-js

Re: JavaScript Physics Engine

#12
post #6
post #2

Cool, how is this diferent from http://box2d-js.sourceforge.net/

In the project, it says box2d-js.sf isn't being updated anymore. Would've been nice to take over the older project so people who were using that one would know there was renewed interest in maintaining this code.

hopefully this one keeps active looks very good

Re: JavaScript Physics Engine

#13

The performance of the demo is awesome. Has anyone tried to do anything more complicated and tested performance. Would be fun to make a js game from this.

TBH it's a small demo - I've heard while JS engines these days are fast, they still struggle with complex physics simulations. I can't see any performance details/improvements noted in this release - anyone got figures?

Untrue, I used a port of Box2d (not this port, our own version) to build an "HTML5" game for sony, see link - http://www.youtube.com/watch?v=AicnuLQxHoQ

Unfortunately the game was taken down due to the Japan situation. Understandable.

Under Safari i was able to simulate ~80 live physics objects in a high collision environment at full frame render, bound to the DOM at 60fps on reasonable hardware 6 months ago (safari still has the fastest DOM paint time out there now).

Even when using Firefox 3.6 i could manage around 20 dynamic objects.

No CSS3 animation, just math and re-paints.

Javascript Math is not slow (at all), its generally the DOM/Canvas slowing you down.

Re: JavaScript Physics Engine

#16

The performance of the demo is awesome. Has anyone tried to do anything more complicated and tested performance. Would be fun to make a js game from this.

TBH it's a small demo - I've heard while JS engines these days are fast, they still struggle with complex physics simulations. I can't see any performance details/improvements noted in this release - anyone got figures?

It's also worth checking out GWTBox2d, which is actually Java code run through GWT, and runs surprisingly well: http://gwtbox2d.appspot.com/

I'm really impressed with what JS can do these days...

Re: JavaScript Physics Engine

#18

I open sourced my "toy" version of JavaScript physics engine just few days ago too. It's not really for serious work, but may be (hopefully) simple and good enough for beginning programmers to learn one way of starting such a project. https://github.com/fuzzthink/SPE.js

It looks good. Have you tested it with SVG? I've been using https://github.com/hrj/box2d-js

Thanks! No, haven't had the time to do an SVG demo, but that shouldn't stop anyone from doing so ;) Thanks for the link, didn't know about jquery.svg, maybe I'll play around with that if I find the time or no one builds one already.

Re: JavaScript Physics Engine

#20

Earlier quoted context omitted.

TBH it's a small demo - I've heard while JS engines these days are fast, they still struggle with complex physics simulations. I can't see any performance details/improvements noted in this release - anyone got figures?

Untrue, I used a port of Box2d (not this port, our own version) to build an "HTML5" game for sony, see link - http://www.youtube.com/watch?v=AicnuLQxHoQ Unfortunately the game was taken down due to the Japan situation. Understandable. Under Safari i was able to simulate ~80 live physics objects in a high collision environment at full frame render, bound to the DOM at 60fps on reasonable hardware 6 months ago (safari…

If you've ever used the C++ Box2D, you'll have been able to create hundreds of objects just fine. 80 still isn't much!
Post reply on HN