[ODE] Mismatching results / non-deterministic behaviour

Jon Watte (ODE) hplus-ode at mindcontrol.org
Mon Sep 18 11:37:14 MST 2006


Note that, if you do interest management at the server (so that not 
every client sees the same set of objects) then big-matrix solvers like 
ODE can't possibly be deterministic between each other. You can only be 
deterministic, in ODE, if everybody has the exact same data.

Cheers,

          / h+


Jean de Largentaye wrote:
> Well there's also something else: different machines have different
> floating-point precisions, and this will also effect determinism. It's
> especially a problem if you're trying to run ODE in a networked
> situation (online game). The solution is to have complete control over
> the precision of your processor's FPU*, which can be problematic if
> you're using DirectX
> (http://opende.sourceforge.net/wiki/index.php/FAQ#I_use_DirectX.2C_and_it_seems_like_ODE_has_switched_to_single-precision.21)
>
> On a side note, is this last bit about DirectX switching to
> single-precision still true with the latest versions? I'm an OpenGL
> guy, so I wouldn't know...
>
> *easier said than done, I know, hopefully Google can help you :)
>
> John
>
> On 9/14/06, Hauser Dietmar <D.Hauser at dion-software.com> wrote:
>   
>> That also explains, why results were the same, when being run on
>> different machines, because the ODE random number was never
>> initialised and therefor using the same predefined seed.
>>     
> _______________________________________________
> ODE mailing list
> ODE at q12.org
> http://q12.org/mailman/listinfo/ode
>
>
>   


More information about the ODE mailing list