Game Development Stack Exchange is a question and answer site for professional and independent game developers. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Has anybody used this library before? What were your experiences? Pros/Cons?

share|improve this question
A quick aside: I have seen many many people get absolutely nowhere with their projects because they kept swapping physics engines. Do not let this happen to you. I recommend picking a reasonable one and sticking with it throughout the project. If you didn't like it, pick another one for the next project. – drxzcl Aug 4 '10 at 7:54
@Ranieri, I've seen this happen to me and others with more than just physics and in more areas than just game engines. My thought was to add a thin layer of abstraction so that if I do decide to change later I can wrap the new library and plug it in. – Anthony Arnold Aug 5 '10 at 3:00
You might like to know about the GangstaWrapper then: – drxzcl Aug 5 '10 at 23:12
Oh yeah, just like the OPAL (open physics abstraction layer). I might do it on my own though; more fun! – Anthony Arnold Aug 6 '10 at 0:03
up vote 3 down vote accepted

Here is a great, high-quality blog post about the reason for Wolfire's switch from ODE to Bullet:

The biggest paragraph is most likely this one:

When I tried dropping four hundred boxes on the Overgrowth "desert fort" scene, Bullet performed the simulation twice as fast as ODE did, even without Bullet's optional performance features like multi-threading and OpenCL. One of the factors in its speed is its use of Sony's newly open-sourced SIMD vector math library, which you can find here if you'd like to use it separately.

share|improve this answer
Wow. Looks like I've been getting excited about ODE for nothing. I think I will dump my idea of using it and take another look at bullet. – Anthony Arnold Aug 4 '10 at 3:55
Bullet documentation is very lousy, but the forum is very active and the developers are very helpfull. – Andreas Aug 6 '10 at 6:57

I've personally never used it, but I think this article gives a good idea of pros/cons:

To summarize the article, it looks like ODE is well documented and stable, but doesn't quite have the cutting edge features and efficiency that some of the more actively developed but less documented alternatives have.

(I have used Bullet and I can vouch for it being a rather nice piece of middleware, though as the article suggests, the documentation is a bit iffy).

share|improve this answer
+1 nice link bullet seems interesting i might take a look at it soon :) – Prix Aug 4 '10 at 5:26

Bullet didn't exist the last time I used ODE (via PyODE, which I will say is a very nice API wrapper) so I can't comment on that. The biggest issue I had with it was stability. It tends to be a very jittery simulation if you don't tune it juuuust right (which takes a lot of poorly documented voodoo). If you use big units that isn't such a big deal, but I made the mistake of using smaller numbers. The joints system is very flexible once you start to think of your system in ODE terms. Also a decent annoyance, it uses a bitfield for collision groups, so you can only have 32 of them.

share|improve this answer

Bullet is derived from ODE. When I first saw ODE, it was a bit too poor, in my perception, after playing a bit with Havok and PhysX. The only reason why I was searching for alternatives was that I needed 64 bit double precision float, for a real-size space multiplayer game. Bullet on the other hand had more coming from the box. It has a great community, and lots of developers working on plugins like something to manage concave shapes (they had 2 alternatives - approximation to convex compound objects and simulation of the original trimesh), CCD implementation, model import, etc. It also had wrappers for some of the 3D engines (don't remember exactly which, might be Ogre3D and XNA). It was very much ahead of ODE, and OpenSource.

This was 2 years ago though, much may have changed. But right then, in the same simulations, nVidia PhysX showed much better performance, and was richer in functionality. So if you plan on using 32 bit single precision float, I'm not sure whether you'll find anything better than PhysX.

share|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.