ODE or IrrCollision

Physic · Post by **Physic** » Mon Sep 05, 2005 6:31 pm

what is better? to use collisions from Irrlicht or use ODE?

Acki · Post by **Acki** » Mon Sep 05, 2005 11:02 pm

I prefer "Newton Games Dynamics" !!!

Guest · Post by **Guest** » Tue Sep 06, 2005 6:52 am

"tokomak"

Guest · Post by **Guest** » Tue Sep 06, 2005 7:22 am

Anonymous wrote:"tokomak"

Tokamak is excluded.

ODE is because of on my possible feature work they use JavaODE. So, I think about it... but also I know, that ODE has problems....

Is ODE dead or we will see 0.6 version?
Also I know(hear) that Newton in next version give opportunity to choose methods between speed and precision.

Guest · Post by **Guest** » Tue Sep 06, 2005 7:42 am

and does anyone know where the forum about ODE? (not mail-list)

Andreas · Post by **Andreas** » Tue Sep 06, 2005 7:49 am

I like the irrlicht collision detection, allthough it's really limited

MACARRAUM · Post by **MACARRAUM** » Tue Sep 06, 2005 4:19 pm

first ODE not dead. for me ODE is the most "professional" quality Physics Lib. And it OpenSource

second, ODE forum: http://ode.petrucci.ch/

Tim · Post by **Tim** » Wed Sep 07, 2005 6:41 pm

MACARRAUM wrote:first ODE not dead. for me ODE is the most "professional" quality Physics Lib. And it OpenSource

second, ODE forum: http://ode.petrucci.ch/

Would you care to elaborate what are the strong points in ODE that makes it the most "professional" quality Physics Lib maybe I overlooked that. Lots of people seem to have the same opinion, but I never found anyone how can answered. For me ODE did not do anything, and I am a PHD in math.

As much as I was attracted by the open source license after month and month of endless frustration, I bought the bullet and switch engine, and as far as I know now ODE source is not worth the aggravation, IMO it could be better.

MACARRAUM · Post by **MACARRAUM** » Wed Sep 07, 2005 7:24 pm

ok ODE is not perfect, but which engine free of physics is perfect? I am working in a commercial project, and we study all current engine for commercial projects (ODE, Newton, Tokamak), and after some tests and discutions, we arrive the opinion that best is the ODE. It possesss some errors, but that they can easily be corrected, case you has certain experience in basic mathematics for games.

Most comercial games using ODE: http://ode.org/users.html

The most impressive game is Xpand Rally.

Now show me other projects using Newton or Tokamak, its fact ODE is most used? Another engines not have comercial projects. But personal or amateur projects.

MACARRAUM · Post by **MACARRAUM** » Wed Sep 07, 2005 7:26 pm

sorry double post sux : /

Tim · Post by **Tim** » Wed Sep 07, 2005 9:09 pm

MACARRAUM wrote:ok ODE is not perfect, but which engine free of physics is perfect? I am working in a commercial project, and we study all current engine for commercial projects (ODE, Newton, Tokamak), and after some tests and discutions, we arrive the opinion that best is the ODE. It possesss some errors, but that they can easily be corrected, case you has certain experience in basic mathematics for games.

You keep repeating ODE it is the best like your prefer taking point, but you do not answer the question. The answer would be: it is the best because if has … <fill in the blank>…..
Instead you just said that you picked a defective solution, based on the hope the you will be able to fix the source. (you know waht they say love is blind)
Well in my experience in several years seeing many qualified engineers wrestling ODE none has subsided, I has seen lots of failure though.

About that game panda, I am not too much into game programming, but it is my experience that you can not under any circurnstance make a stable vehicle with ODE, because the joints are just too soft and in not way can be fixed. I can assure you that, that game may be using ODE, but it is not for car physics, car physics was done the same way it is done for any other car game; a box mounted in four springs.
For your information car physics and fast speed joints is not something physics engines are good at. ODE in particular is the worst at it.
And about that list of titles, I think you should check the list again, you will find that in about 80 % to 90% has switched to nobodex or other free engine.

Hey If you are happy with ODE, then that’s very good, more power to you, I was just curios as if you has a reason to make such a strong statement as ODE it is the best. without any objective proof.

Tim · Post by **Tim** » Wed Sep 07, 2005 9:22 pm

MACARRAUM wrote:ok ODE is not perfect, but which engine free of physics is perfect? I am working in a commercial project, and we study all current engine for commercial projects (ODE, Newton, Tokamak), and after some tests and discutions, we arrive the opinion that best is the ODE. It possesss some errors, but that they can easily be corrected, case you has certain experience in basic mathematics for games.

You keep repeating ODE it is the best like your prefer taking point, but you do not answer the question. The answer would be: it is the best because if has … <fill in the blank>…..
Instead you just said that you picked a defective solution, based on the hope the you will be able to fix the source. (you know waht they say love is blind)
Well in my experience in several years seeing many qualified engineers wrestling ODE none has subsided, I has seen lots of failure though.

About that game panda, I am not too much into game programming, but it is my experience that you can not under any circurnstance make a stable vehicle with ODE, because the joints are just too soft and in not way can be fixed. I can assure you that, that game may be using ODE, but it is not for car physics, car physics was done the same way it is done for any other car game; a box mounted in four springs.
For your information car physics and fast speed joints is not something physics engines are good at. ODE in particular is the worst at it.
And about that list of titles, I think you should check the list again, you will find that in about 80 % to 90% has switched to nobodex or other free engine.

Hey If you are happy with ODE, then that’s very good, more power to you, I was just curios as if you has a reason to make such a strong statement as ODE it is the best. without any objective proof.

Guest · Post by **Guest** » Wed Sep 07, 2005 9:26 pm

sorry double post.

Anyway I think all of the engine are good for. I am using Gangta that allows my experiment more than one engine at once.

Guest · Post by **Guest** » Thu Sep 08, 2005 7:12 pm

hm.

So, is it better AND EASIER to use Irrlicht TriMesh collisions?

I dont plan to use physic scenes with forces and so on... just collision detection is my aim for near feature.

And I think that Niko will improve it in irrlicht. (?)

Tim · Post by **Tim** » Fri Sep 09, 2005 9:29 pm

It is easier probably but not better.
Collision is some very complex subject that it is not the strong point of graphics engines.
I you just want simple collision you can probably try opcode,
It is easy and open source.
It is not as powerful as using a physics engine but it would be more efficient that using Irrlicht collision