[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Public WebGL] cWebGL, webgl implemented in JavaScript



I wouldn't completely discount the idea
of reasonably fast JavaScript software 3d engines. For example:
http://www.zynaps.com/site/experiments/environment.html?mesh=teapot.wft
. On firefox 3.6.25 I get around 2fps, but on Firefox 9 I get a solid
50fps (on a Core 2 duo E8400). There's also a Quake 3 level demo on that site too (was about 10fps on FF8, seems to be broken in FF9 though). Obviously those kinds of performance
boosts can't go on indefinitely, but there is still scope for
improvement.


There's also plenty of scope for fairly
decent software 3d engines generally. Take Stage3, for example. If
you go to http://helloracer.com/ you can do a side by side comparison
of Flash and WebGL rendering of an F1 model. If I go to the WebGL
version, it initially won't work (I'm on Ubuntu 10.10 (Mesa
9.something)), so I have to force enable WebGL. When I do I get a
nice smooth framerate. I also get rendering bugs and about
100% CPU usage (no layers acceleration). If I go to the Flash version
I get perfectly smooth rendering, no rendering bugs, and about 200% CPU usage. It's
maybe sightly unfair to suggest that WebGL GPU rendering on my system
is only twice as fast as CPU rendering, but even if I did have the
most up to date graphics drivers I would still not get layers
acceleration by default, and so would still have the same performance issues. 

Liam


----- Original Message -----
From: Benoit Jacob <bjacob@mozilla.com>
To: Foo Bar <cosinusoidally@yahoo.co.uk>
Cc: public webgl <public_webgl@khronos.org>
Sent: Sunday, 8 January 2012, 0:01
Subject: Re: [Public WebGL] cWebGL, webgl implemented in JavaScript


This already is an amazing achievement, and I would be interested in how well it does on the WebGL conformance test suite.

But realistically, this can't offer practical performance levels outside of very simple scenes. It's a great technology prowess and might even be adequate for some niche real-world uses, but don't bet on it as a realistic avenue to get WebGL supported everywhere.

Regarding a 'reference implementation', this isn't really how web standards work. Instead, we aim to have multiple implementations fully passing the conformance test suite.

Cheers,
Benoit
----- Original Message -----
> 
> 
> 
> Hi
> 
> Thought you lot might be interested in cwebgl
> http://code.google.com/p/cwebgl/ . It's a webgl implementation
> written in JavaScript. It works in Firefox, IE 9 and Chrome. There's
> a bunch of demos on: http://cimaron.net/projects/cwebgl/ .
> 
> 
> I don't think it is anywhere near complete yet, but the author has
> done some great work. It could even be used as a the basis for a
> much needed reference webgl implementation?
> 
> 
> Liam Wilson
> 
> 
> -----------------------------------------------------------
> You are currently subscribed to public_webgl@khronos.org.
> To unsubscribe, send an email to majordomo@khronos.org with
> the following command in the body of your email:
> unsubscribe public_webgl
> -----------------------------------------------------------
> 
> 

-----------------------------------------------------------
You are currently subscribed to public_webgl@khronos.org.
To unsubscribe, send an email to majordomo@khronos.org with
the following command in the body of your email:
unsubscribe public_webgl
-----------------------------------------------------------

-----------------------------------------------------------
You are currently subscribed to public_webgl@khronos.org.
To unsubscribe, send an email to majordomo@khronos.org with
the following command in the body of your email:
unsubscribe public_webgl
-----------------------------------------------------------