[ODE] PATCH: flat-capped cylinder support

Bram Stolk bram at sara.nl
Fri Nov 11 10:07:15 MST 2005


Olivier Michel wrote:
> Hello,
> 
> This sounds great. I have been using the dCylinder contrib and found it 
> a bit buggy (like collision with the caps are not always detected). What 
> is annoying with the dCylinder2 is that there is no Cyl/Cyl collision 
> detection, right ?
> Otherwise, does it work fine for you ? Did anyone have a chance to 
> compare dCylinder2 to dCylinder contribs (in term of stability, bugs, 
> completness, etc.) ? I am considering switching from dCylinder to 
> dCylinder2, but would like to share opinions about it before...
> 

dcyl2 has indeed no cyl vs cyl, only box,sphere,mesh.
it has one bug, which does not affect me: cyl2 intersection
with a triangle planar to cylinder-cap is not detected.
This is easy to avoid, as co-planarity is only achieved
when explicitly set as startup condition, and not likely to be reached
by the sim.

I find it to have excellent suitability for wheel vs terrain.
Even a car on its side has no 100% planar wheel/ground, so it
works just fine.

I'll post some results with dcyl2 in my game, later.

   Bram

> Cheers,
> 
> -Olivier
> 
> Bram Stolk wrote:
> 
>> Hello!
>>
>>
>> I've added flat-capped cylinder support to my working copy of the 
>> unstable
>> branch. (based on dcyl2 contrib).
>>
>> It can be enabled with a configure option:
>>
>> $ ./configure --with-cylinder
>>
>> Default is no flat-capped cyl support.
>>
>>
>> I've tested both builds (with and without) on linux with g++-4 and
>> it works just fine.
>>
>>
>> It would be great if I could get cvs-write access to the unstable
>> branch, so I can commit. Could this be arranged?
>>
>> If not, what is the proper way to contribute this?
>> I've put the output of 'cvs diff' at
>> http://stolk.org/tmp/ode/
>> This dir also contains 3 new files, to go in ode/src
>>
>> Also: I think someone erroneously committed the top level
>> Makefile to the CVS repo. I think this should be deleted,
>> as Makefile is derived indirectly from Makefile.am which is
>> already in CVS.
>>
>>   Bram Stolk
>> _______________________________________________
>> ODE mailing list
>> ODE at q12.org
>> http://q12.org/mailman/listinfo/ode
>>
>>
> _______________________________________________
> ODE mailing list
> ODE at q12.org
> http://q12.org/mailman/listinfo/ode




More information about the ODE mailing list