[ODE] what is the relationship between Croteam and ODE?

[Russ Smith]

even the time to collide the simplest primitives together is greater
than the time for one indirect call + cache line load, so i wouldn't be
worrying about that dispatch table (unless all you're doing is colliding
spheres together and you need to squeeze out every last cycle).

russ.

--
Russ Smith
http://www.q12.org/

On Mon, 5 Apr 2004, Jon Watte wrote:

>
> > on the issue of overhead of adding more collision primitives, there is
> > none: collision functions are called by looking up function pointers in
> > an N*N table.  more collision primitives just makes the library bigger.
>
> Is the size of that table, and the number of cache lines used by it, of significance to run-time performance, or are we assuming that the LCP solver pretty much makes the rest of the system "noise" ?
>
> Cheers,
>
> 			/ h+
>
>
>