Re[2]: [ODE] tri-callider

[slipch]

Hello Erwin,
   I look forward your new version (I need it yesterday indeed).
   In the old one my boxes penetrate into mesh in some cases
   especially when pushing them into by some forces.
   
   As to odd contact points in my version there may be two reasons for
   boxes. First, I used too many axes for separation it seems that some of them can be removed
   Another reason may be points which adding for stability I add them
   only for one case - separation along triangle normal but do not
   check if they are in the triangle boundaries I simply copy a pies
   of code from ODE Box- Plain collider. But I do not think that it
   possible to notice it in simulation because the points adding only
   when box side quasi parallel to the triangle plain.

Thursday, August 15, 2002, 12:46:22 PM, you wrote:

EdV> This is interesting. I've just tested ran it through my (quick) tests, and
EdV> it indeed works much better than the old version. I dont think its as good
EdV> as my new tricollider (which will be released later this week) though, but
EdV> that may also depend on what you're using it for.

EdV> Sometimes it seems to generate odd contact points, which is probably the
EdV> reason for some instabilities. Anyways, this is a very good effort.

EdV> Erwin

EdV> ----- Original Message -----
EdV> From: "slipch" <slipch@gsc-game.kiev.ua>
EdV> To: <ode@q12.org>
EdV> Sent: Wednesday, August 14, 2002 10:36
EdV> Subject: [ODE] tri-callider




EdV> _______________________________________________
EdV> ODE mailing list
EdV> ODE@q12.org
EdV> http://q12.org/mailman/listinfo/ode



-- 
Best regards,
 Konstantin Slipchenko                         mailto:slipch@gsc-game.kiev.ua