rigid body physics in Blender (was Re: [Bf-committers] SOC idea)

Ed Halley ed at halley.cc
Mon May 1 01:26:51 CEST 2006


Has anyone been looking at OPAL for providind a higher-level rich 
interface to various physics?  OPAL is currently a wrapper over ODE but 
it's intended as a general wrapper that allows any number of physics 
solvers to operate as drop-in replacements or collaboratively.  This 
could then insulate the scripters from caring about the specifics of ODE 
or BULLET or SUMO or other constraint solvers, and just code to one 
common API instead.

http://opal.sourceforge.net/  --  Open Physics Abstraction Layer

I haven't used it personally yet, but I will be using a (commercial) 
toolkit that has selected OPAL for their python physics modeling 
mechanisms, and thought I'd raise it here for the list.

Willian Padovani Germano wrote:
> Hi Erwin,
> 
> Should have emailed you already about this. You probably know that 
> integrating ODE in Blender has been in the plans for some time (and 
> available in experiments via Python probably for longer than that).
> 
> I've made some quick tests a couple months ago, been reading ODE mailing 
> list archives and the BULLET forum and now and then we discuss a little 
> about it during irc meetings. Now I should have time to try to code it 
> properly, using groups to define spaces, possibly adding a new type in 
> Blender to represent joints, etc.
> 
>  From what I've gathered (specially from your posts to the mentioned 
> list and forum), it seems it would be desirable to have both ODE and 
> BULLET integrated in Blender, at least with BULLET as alternative for 
> ODE's  collision detection (you released ODE with BULLET support some 
> months ago, from what I recall). We can also have both as alternatives 
> for dynamics, you're obviously the right person to opinate about that.
> 
> So, in short:
> 
> - Integrating rigid body dynamics in Blender directly is a current and 
> "approved" project.
> - There are coders interested, experimenting in C or Python with ODE.
> - A C API for BULLET is a welcome addition, for sure.
> - The details have not been decided yet: if rigid body should become an 
> object constraint (Ton's informal suggestion), modifier or be independent.
> - I'd like to know your opinions about the integration: ODE + BULLET?
> 
> As I said I'm interested and working on the ODE side and it's great to 
> see you interested in working on physics in Blender, too. Unless you 
> prefer to code it all yourself "in one go", I and probably others here 
> would enjoy participating.
> 
> Finally, as Ton frequently points, properly integrating rigid body 
> physics with particles,  softbodies, cloth, and the rest of the 
> animation system is also an important goal.
> 
> This seems to be a good time to discuss / coordinate efforts, so others 
> interested please say so, too.
> 
> Erwin Coumans wrote:
> 
>> In that case it would be useful to integrate Bullet collision 
>> detection / physics more tightly into Blender (independent from the 
>> Game engine).
>> With some performance improvements it would be very useful for 
>> 'advanced' particles too. I can easily add a C-API to it (to avoid C++ 
>> hassle).
>>
>> I would like to do this, but need someone who helps pointing out the 
>> hooks in Blender. Would it be a rigidbody 'modifier'?
>> Erwin
> 
> 
> -- 
> Willian
> _______________________________________________
> Bf-committers mailing list
> Bf-committers at projects.blender.org
> http://projects.blender.org/mailman/listinfo/bf-committers
> 
> 
> 

-- 
[ e d @ h a l l e y . c c ]


More information about the Bf-committers mailing list