Hi,

there is the Physics Abstraction Library
(http://www.adrianboeing.com/pal/index.html ,
http://pal.sourceforge.net). I'm not sure to what extend that is
useful, but it seems to be rather popular. It already abstracts ODE,
Bullet, and other, even commercial, engines. Possibly worth a look.

I actually plan to use OSG (or Ogre3D, the jury is still out) with PAL
for a robotics simulator, which might become the successor of USARSim
(http://usarsim.sourceforge.net). Some more direct integration between
OSG and some physics abstraction (PAL or homegrown) would be great!

Cheers,
max

On Tue, Nov 4, 2008 at 10:27 AM, Robert Osfield
<[EMAIL PROTECTED]> wrote:
> Hi Guys,
>
> It's good to see a number of different users popping with discussion
> of their own work on integrating physics with the OSG.  I would be
> great to have an OSG NodeKit for doing physics, but I'm very wary of
> tying to one specific engine as each engine has it's own strengths and
> weakness, and I also don't want to steer people away from making
> particular physics engine choices including closed source ones - such
> as Vortex, users should be able to integrate the best tool for they
> job they can afford.
>
> So I'd like to punt the possibility of having some kind of base
> osgPhyics API that makes it easier to bolt on different physics
> engines onto the OSG.   This would require spotting the commonality
> between the different engines and distilling this, for instance the
> osgViewer library does this trick to a certain extent with the
> different GraphicsWindow implementations.  You also don't want to hide
> too much of the implementation advanced phyics engine tuning will
> probably require users to grapple with the lower level physics engines
> settings.  While abstracting completely is impracticable it'd be good
> to provide the template for the concrete implementations, and also
> make it easier for applications to move between different physic
> engines at the backend.
>
> Is such a lib possible or practical?  It'll be more wore initially
> than just an osgODE, osgBullet or osgVortex library but longer term it
> could reduce the cost of maintenance of all these different
> variations.   The first step in this direction would be to see the
> code of various  physic engine + OSG integration with a view to
> spotting the commonality between them.
>
> Thoughts?
> Robert.
> _______________________________________________
> osg-users mailing list
> [email protected]
> http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org
>
_______________________________________________
osg-users mailing list
[email protected]
http://lists.openscenegraph.org/listinfo.cgi/osg-users-openscenegraph.org

Reply via email to