[ODE] Double precision patch

[Bill Sellers]

Hi,

I just installed the unstable branch of ODE and I had to make the  
following patch to collision_std.cpp to get it to compile as double  
precision (it's currently hardwired as float * rather than dReal *):

diff -r1.19.2.3 collision_std.cpp
143c143
<   const float* pos = s->final_posr->pos;
---
 >   const dReal* pos = s->final_posr->pos;
338,339c338,339
<   const float* R = g->final_posr->R;
<   const float* pos = g->final_posr->pos;
---
 >   const dReal* R = g->final_posr->R;
 >   const dReal* pos = g->final_posr->pos;

Looking around ode/src I notice there are quite a few other hardcoded  
floats casts that probably ought to be dReals - although at the  
moment all they will do if affect precision. They are all in  
collision_* routines.

Cheers
Bill

-- 
Dr. Bill Sellers                            Email:  
William.Sellers at manchester.ac.uk
Lecturer in Integrative Vertebrate Biology
Faculty of Life Sciences                    Tel.   0161 2751719
3.614 Stopford Building                     Fax:   0161 2753938
The University of Manchester                Mob:   0796 0024359
Oxford Road, Manchester, M13 9PT, UK        http:// 
sellers.sbc.man.ac.uk/ASL/