Question about the improved collision detection and response

Discussions Related to Game Engines Aside from L. Spiro Engine

Question about the improved collision detection and response

Postby Ivan » Sun Jul 01, 2012 5:02 am

Hi, L. Spiro, I read the improved collision detection and response article and implemented it, I come here to ask a question, because a thread you post on gamedev.net(http://www.gamedev.net/topic/618036-improved-collision-detection-paper-question/), it seems that you read the article and may implement it yourself. In my own implementation, it sometimes got stuck, changing the verycloseDistance can reduce the times got stuck, but it got stuck sometimes. So my question is, is it a right method without flaws or there is something wrong in my code? I tried to contact the author, but there is no reply. So, can you give me some suggestions? Thanks in advance. :)
Ivan
I Have a Question
 
Posts: 3
Joined: Sun Jul 01, 2012 4:49 am

Re: Question about the improved collision detection and resp

Postby L. Spiro » Sun Jul 01, 2012 6:35 am

I have not implemented the routine described by that paper, but this is a problem common to all collision systems.

However you will need to explain what you mean by “stuck” and what you are trying to do, the architecture of your engine, etc.
Physics should be handled at fixed-step intervals, for example.


L. Spiro
It is amazing how often people try to be unique, and yet they are always trying to make others be like them.
- L. Spiro 2011
L. Spiro
Site Admin
 
Posts: 54
Joined: Thu Jul 21, 2011 2:59 pm
Location: Tokyo, Japan

Re: Question about the improved collision detection and resp

Postby Ivan » Sun Jul 01, 2012 7:18 am

L. Spiro wrote:I have not implemented the routine described by that paper, but this is a problem common to all collision systems.

However you will need to explain what you mean by “stuck” and what you are trying to do, the architecture of your engine, etc.
Physics should be handled at fixed-step intervals, for example.


L. Spiro

By stuck, I mean when the player walks on the terrain, it stops moving somewhere on the terrain, and can't move any more. What I want is to let the player collide with the terrain and static objects in the game world, a simple collision system, not the rigid body and soft body dynamics etc. Any suggestions to solve the 'stuck' problem, or other methods that other engines use. The scene is managed by quadtree or octree, and the bounding volume is aabb. What do you mean 'a problem common to all collision system', how do they work it out?
Ivan
I Have a Question
 
Posts: 3
Joined: Sun Jul 01, 2012 4:49 am

Re: Question about the improved collision detection and resp

Postby L. Spiro » Sun Jul 01, 2012 9:35 am

It means you sink into the terrain and you are not pushing yourself out.
You need to determine how deep you are into the terrain and send an impulse to kick yourself out.

You can find information on how to do this via the keywords, “Resolving Interpenetration”, “Interpenetration Linear Projection”, etc.


L. Spiro
It is amazing how often people try to be unique, and yet they are always trying to make others be like them.
- L. Spiro 2011
L. Spiro
Site Admin
 
Posts: 54
Joined: Thu Jul 21, 2011 2:59 pm
Location: Tokyo, Japan

Re: Question about the improved collision detection and resp

Postby Ivan » Sun Jul 01, 2012 10:26 am

Thanks a lot, I will search the key words and try to solve the problem.Thanks. :)
Ivan.
Ivan
I Have a Question
 
Posts: 3
Joined: Sun Jul 01, 2012 4:49 am


Return to Other Game Engines

Who is online

Users browsing this forum: No registered users and 0 guests

cron