[Opensim-dev] Super easy question

[Cristina Videira Lopes]

yes, that is the missing "(*) restrictions apply" on what I said :)
Bad for whoever wants to add a new physics engine, but ultimately good  
for OpenSim. Making the physics API more robust to specific engines  
can only be a good thing.

On Aug 20, 2010, at 1:55 PM, Hurliman, John wrote:

> Actually, if you actually try to implement Havok (or Bullet or  
> PhysX) in OpenSim you will come up against a few walls. The physics  
> API in OpenSim is highly tailored for the specific way ODE is being  
> used right now, and many ways of implementing physics including the  
> "standard" (or at least the most documented and tested) ways in  
> Havok/Bullet/PhysX simply will not work. You would have to rewrite  
> the physics interface in OpenSim to give the individual  
> implementations more freedom in how they do the simulation. I'm not  
> saying it's impossible, in fact there have been at least two  
> attempts already to add Bullet physics to OpenSim, but you either  
> have to figure out how to make your physics engine of choice behave  
> exactly like the way ODE is being used right now or redo the physics  
> API in OpenSim.
>
> John
>
>> -----Original Message-----
>> From: opensim-dev-bounces at lists.berlios.de [mailto:opensim-dev-
>> bounces at lists.berlios.de] On Behalf Of Cristina Videira Lopes
>> Sent: Friday, August 20, 2010 1:50 PM
>> To: opensim-dev at lists.berlios.de
>> Subject: Re: [Opensim-dev] Super easy question
>>
>> I'm just explaining the mechanics of the process. If Bri or anyone
>> else here has enough time and/or resources to do that kind of
>> integration, by all means do it.  Just because Havok is a no go for
>> the stock distribution of OpenSim doesn't mean that that kind of work
>> should be discouraged.
>>
>> Technically, OpenSim is ready for it because of its highly extensible
>> architecture regarding just about every important component of the
>> system. But clearly, even in the best case scenario of being able to
>> use the binaries as they come, this is still not a plug-and-play  
>> deal;
>> there's a c#/.NET layer that needs to be written.
>>
>>
>> On Aug 20, 2010, at 1:19 PM, Robert A. Knop Jr. wrote:
>>
>>> On Fri, Aug 20, 2010 at 01:11:32PM -0700, Michael Cerquoni wrote:
>>>> I think it should also be mentioned that just because Second Life
>>>> is using
>>>> Havok, they are not using standard binaries that are freely
>>>> available, they
>>>> have no doubt highly customized the engine to meet their needs.
>>>
>>> Indeed.  I don't know exactly what was done, so I'm not giving away
>>> inside information here.  But do remember that the project to
>>> integrate
>>> Havok 4 into SL (upgrading from Havok 2) took many months, first of
>>> development, then of testing.  At least one Linden (Andrew) was
>>> working
>>> near-full-time on the Havok 4 project for quite some time.  It's not
>> a
>>> plug-and-play situation.
>>>
>>> --
>>> --Rob Knop
>>> E-mail:    rknop at pobox.com
>>> Home Page: http://www.pobox.com/~rknop/
>>> Blog:      http://scientopia.org/blogs/galacticinteractions/
>>> _______________________________________________
>>> Opensim-dev mailing list
>>> Opensim-dev at lists.berlios.de
>>> https://lists.berlios.de/mailman/listinfo/opensim-dev
>>
>> _______________________________________________
>> Opensim-dev mailing list
>> Opensim-dev at lists.berlios.de
>> https://lists.berlios.de/mailman/listinfo/opensim-dev
> _______________________________________________
> Opensim-dev mailing list
> Opensim-dev at lists.berlios.de
> https://lists.berlios.de/mailman/listinfo/opensim-dev