[ODE] Sweep-and-Prune space for ODE

Brian Marshall brianmarshall at blueyonder.co.uk
Thu Nov 25 17:23:58 MST 2004

Which implementation of std:sort was it? Might be worth applying the same
approach - since ODE's in C, it can't just be dropped in. It would be useful
in removing the float/double problem anyway.

It'd be interesting to see how the performance goes between the different
sorts in a complex test. I do wonder if the worse case behaviour for some
sorts wouldn't come into play with sweep and prune. Certainly in contrived
box stacks, lots of identical values aren't ideal for some sorts, while
radix doesn't care.

That's really going to be the case with the different broad phase collision
systems anyway - which is 'best' is going to rely quite a lot on the data.


-----Original Message-----
From: Graham Fyffe [mailto:gfyffe at gmail.com] 
Sent: 25 November 2004 17:10
To: Brian Marshall
Cc: ode
Subject: Re: [ODE] Sweep-and-Prune space for ODE

Oops, I forgot to Reply All :P  Here it is again:

This is exactly the source code I tried that was 2.5 times slower than
std::sort.  This was pretty consistent over all the sizes I tested.

- Graham

More information about the ODE mailing list