News for hackers. Stuff that matters.
Re:Three.js/Ammo.js Quake 3 Physics Test by experiment0 on Monday April 09, 2012 @06:44AM (#3816425)
Impressive but the controls really let it down. Moving around made me feel sea sick with the room spinning everywhere. The cursor needs to be fixed to the centre of the screen (as in traditional fps's) and not have some sort of momentum whereby you continue spinning round when you're facing the direction you want.

Aside from that it looks really nice!

Re:Three.js/Ammo.js Quake 3 Physics Test by keyle on Monday April 09, 2012 @07:27AM (#3816513)
I think you're missing the point. It's a demo of the strength of WebGL and a milestones towards 3D in the browser.
Re:Three.js/Ammo.js Quake 3 Physics Test by experiment0 on Monday April 09, 2012 @07:37AM (#3816542)
No I get that and like I said I think its really impressive! But, I think that the controls detract from properly delivering the message (and it would be much more enjoyable to use if movement was less fiddly).
Re:Three.js/Ammo.js Quake 3 Physics Test by reitzensteinm on Monday April 09, 2012 @10:24AM (#3817063)
I completely agree. I'm a game developer, and I'd have a hard time intentionally designing a worse mouse look. Ouch. (I realize that they're not using mouse capture intentionally).

That said, it's a fantastic demonstration of what's possible in the browser. I'm really looking forward to seeing what three.js is like in a year or two, and how that rivals Flash.

Re:Three.js/Ammo.js Quake 3 Physics Test by augustl on Monday April 09, 2012 @08:05AM (#3816605)
AFAIK it's not possible to capture the mouse like that in browsers and/or WebGL.
Re:Three.js/Ammo.js Quake 3 Physics Test by daeken on Monday April 09, 2012 @08:13AM (#3816624)
Actually, pointer lock is supported in Chrome and Firefox at the least. It still needs work, but it solves most of the key issues.
Re:Three.js/Ammo.js Quake 3 Physics Test by kodisha on Monday April 09, 2012 @07:05AM (#3816467)
THREE.PerspectiveCamera( 50, window.innerWidth / window.innerHeight, 1, 20000 );

Why on earth did you use FOV of 50? Default on quake is 90, and in most games.

Also, 95% of the professional players play with fov in range 100 - 130. (I personally play with 130)

For the sake of the demo, i think this needs to be fixed to at least 90 deg. This is like trying to walk trough room looking trough a plastic pipe, you see whats in front of you, but you need way more information to move comfortably.

Re:Three.js/Ammo.js Quake 3 Physics Test by jlouis on Monday April 09, 2012 @10:14AM (#3817020)
Interestingly, among the professional quake live players, the average fov seems to be in the ballpark of 90 to 100. Most of the aim-heavy players with a 90 fov.

My guess is that players who play team-based games might like to have larger fovs, but 130 is definitely a bad choice from the perspective of hitting an opponent.

The point about consoles using a 50 fov is interesting though. It kinda makes sense since most console FPS games are just point-n-shoot-hide-n-seek types of games.

Re:Three.js/Ammo.js Quake 3 Physics Test by talmand on Monday April 09, 2012 @09:44AM (#3816900)
It's a trick used by consoles to increase performance.
Re:Three.js/Ammo.js Quake 3 Physics Test by richbradshaw on Monday April 09, 2012 @09:57AM (#3816951)
And because they are typically further away, so the view cone needs to be smaller feel right.
Re:Three.js/Ammo.js Quake 3 Physics Test by rangibaby on Monday April 09, 2012 @07:10AM (#3816480)
I agree with post. Around 50 seems to be the default for console shooters these days - which is a shame when they get ported to PC with a locked POV :-|
Re:Three.js/Ammo.js Quake 3 Physics Test by NiekvdMaas on Monday April 09, 2012 @07:43AM (#3816552)
It took me a while to figure out what the physics part of this demo was. You have to elevate up a bit (click to move) and then throw boxes using the spacebar. Very nicely done by the way!
Re:Three.js/Ammo.js Quake 3 Physics Test by itsnotlupus on Monday April 09, 2012 @07:50AM (#3816571)
Use ASWD to move, mouse around to get motion sickness, SPACE to throw a cube around.
Re:Three.js/Ammo.js Quake 3 Physics Test by ck2 on Monday April 09, 2012 @09:20AM (#3816831)
What's amazing is I do NOT have hardware accelerated video yet Firefox 12 is not even breaking a sweat.

My CPU temps are only a couple degrees higher and it's very smooth.

Re:Three.js/Ammo.js Quake 3 Physics Test by torbenc on Monday April 09, 2012 @07:28AM (#3816514)
I saw the FPS graph and was about to close it, because I thought there was nothing more happening. Add a "Map loading" indicator or something like that.
Re:Three.js/Ammo.js Quake 3 Physics Test by lucaspiller on Monday April 09, 2012 @07:50AM (#3816573)
Yeah Three.js is a bit strange like that. It loads all textures asynchronously but doesn't provide a (built in) way to have any kind of loading indicator, so for a lot of the demos until they are fully loaded you just see a blank screen. The car demos are a great example of this: http://mrdoob.github.com/three.js/examples/webgl_materials_c...
Re:Three.js/Ammo.js Quake 3 Physics Test by chib on Monday April 09, 2012 @10:03AM (#3816981)
Fairly impressive how many boxes you can spawn before it starts to stutter.
Re:Three.js/Ammo.js Quake 3 Physics Test by bigdubs on Monday April 09, 2012 @12:28PM (#3817666)
awesome demo! interesting to see the progress.

one thing that is very noticeable though; no multi-threading! as soon as you start moving the 'cube' stops. my guess is physics updating defers to player movement.

Re:Three.js/Ammo.js Quake 3 Physics Test by Lewton on Monday April 09, 2012 @08:17AM (#3816645)
Holding down space for a couple of seconds was fun!
Re:Three.js/Ammo.js Quake 3 Physics Test by nknight on Monday April 09, 2012 @11:53AM (#3817503)
Fun while it lasted. Flew around a bit, was very impressed, then looked toward a teleporter and froze my whole machine. Presumably a Google or Apple/NVIDIA bug (Chrome 18.0.1025.151 on a mid-2010 MBP with a GeForce GT 330M and fully-updated Lion).

Disinclined to try again since I've got work to do. :P