You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@ode.apache.org by chris <dr...@gmail.com> on 2007/01/04 14:25:12 UTC
ODE sensitivity to displacement in space or time
Hi,
For my Thesis, I had reason to perform a number of identical
experiments with simple variations: displacements in space or time.
One set of experiments was using ODE:
I did a simple rigid body physics experiment using pyode: dropped a
box near a stationary box onto a plane. I saved an image of the rest
position of the box. The experiment is repeatable: the box lands in
same place each time relative to the stationary one.
Then I moved the whole experiment by (10,10,10) and repeated it. The
box rest position was different.
I also repeated the first experiment with the only difference being
that I programmed it to wait 8000 seconds before dropping the box.
Again the rest position was difference.
What do the developers of ODE have to say as to the cause of the
sensitivity of ODE to changes in position or time?
Attached are images of the rest position in the first and second
experiments and the code used to produce each. Also image of the time
displaced version and its code.
The hardware I was using was a VAIO notebook: intel 770 CPU with
radeon x600 GPU. For python install I followed the steps from Miriam
English:
1. installed python 2.4
python-2.4.3.msi
http://www.python.org/ftp/python/2.4.3/
2. edited autoexec.bat to add:
SET PATH=%PATH%;C:\python
SET PYTHON=C:\PYTHON\
3. installed pyOpenGL
PyOpenGL-2.0.2.01.py2.4-numpy23.exe
http://pyopengl.sourceforge.net
4. Copied glut32.dll to the C:\python\Lib\site-packages\OpenGL
directory. (This is a *crucial* step.)
glut32.dll
http://www.xmission.com/%7Enate/glut.html
5. installed OpenGLContext
OpenGLContext-2.0.0c1.win32-py2.4.exe
http://pyopengl.sourceforge.net/context/
(downloaded from http://pyopengl.sourceforge.net/ )
6. installed PIL (python Image Library)
PIL-1.1.5.win32-py2.4.exe
http://www.pythonware.com/products/pil/
7. installed pyODE
PyODE-1.1.0.win32-py2.4.exe
http://pyode.sourceforge.net/
Now I can double-click on the tutorial3.py from
http://pyode.sourceforge.net/ and it simply runs!
thanks,
chris
Re: ODE sensitivity to displacement in space or time
Posted by chris <dr...@gmail.com>.
Anyway, nice talkin to u guys, I better get off this list now.
I'll sign off with a little process (not to do) for the process heads here :)
1. Wait till late at night
2. Drink some Baileys
3. That went down nice and smooth, drink some more!
4. Find that list you've been meaning to post to.
5. Hmm - that site's down - find where it moved to, ok, seems to be on
this site...
6. Subscribe and post ... Ooops!
cheers,
chris
Re: ODE sensitivity to displacement in space or time
Posted by chris <dr...@gmail.com>.
That's interesting - I wrote something similar once, I think I'll
resurrect it :)
As for MUP - from what I just read - it is about any two "builds" of
Maven will not produce the same result :).
chris
On 1/5/07, Maciej Szefler <mb...@intalio.com> wrote:
> I believe you intended to post this message to the Maven user's
> mailing list. The phenomemon you are observing is consistent with MUP
> (the maven uncertainty principle) which states that two executions in
> different locations or at different times will produce different
> results almost surely. There is a conjecture that states that this
> principle applies to anything built with maven and anything that
> references or is referenced by anything built with maven (directly
> /or/ through transitive dependencies) . While the conjecture is
> generally believed to be true no proof is known at this time.
> -mbs
>
>
>
Re: ODE sensitivity to displacement in space or time
Posted by Maciej Szefler <mb...@intalio.com>.
I believe you intended to post this message to the Maven user's
mailing list. The phenomemon you are observing is consistent with MUP
(the maven uncertainty principle) which states that two executions in
different locations or at different times will produce different
results almost surely. There is a conjecture that states that this
principle applies to anything built with maven and anything that
references or is referenced by anything built with maven (directly
/or/ through transitive dependencies) . While the conjecture is
generally believed to be true no proof is known at this time.
-mbs
On 1/4/07, chris <dr...@gmail.com> wrote:
> Hi,
>
> For my Thesis, I had reason to perform a number of identical
> experiments with simple variations: displacements in space or time.
> One set of experiments was using ODE:
>
> I did a simple rigid body physics experiment using pyode: dropped a
> box near a stationary box onto a plane. I saved an image of the rest
> position of the box. The experiment is repeatable: the box lands in
> same place each time relative to the stationary one.
>
> Then I moved the whole experiment by (10,10,10) and repeated it. The
> box rest position was different.
>
> I also repeated the first experiment with the only difference being
> that I programmed it to wait 8000 seconds before dropping the box.
> Again the rest position was difference.
>
> What do the developers of ODE have to say as to the cause of the
> sensitivity of ODE to changes in position or time?
>
> Attached are images of the rest position in the first and second
> experiments and the code used to produce each. Also image of the time
> displaced version and its code.
>
> The hardware I was using was a VAIO notebook: intel 770 CPU with
> radeon x600 GPU. For python install I followed the steps from Miriam
> English:
>
> 1. installed python 2.4
> python-2.4.3.msi
> http://www.python.org/ftp/python/2.4.3/
>
> 2. edited autoexec.bat to add:
> SET PATH=%PATH%;C:\python
> SET PYTHON=C:\PYTHON\
>
> 3. installed pyOpenGL
> PyOpenGL-2.0.2.01.py2.4-numpy23.exe
> http://pyopengl.sourceforge.net
>
> 4. Copied glut32.dll to the C:\python\Lib\site-packages\OpenGL
> directory. (This is a *crucial* step.)
> glut32.dll
> http://www.xmission.com/%7Enate/glut.html
>
> 5. installed OpenGLContext
> OpenGLContext-2.0.0c1.win32-py2.4.exe
> http://pyopengl.sourceforge.net/context/
> (downloaded from http://pyopengl.sourceforge.net/ )
>
> 6. installed PIL (python Image Library)
> PIL-1.1.5.win32-py2.4.exe
> http://www.pythonware.com/products/pil/
>
> 7. installed pyODE
> PyODE-1.1.0.win32-py2.4.exe
> http://pyode.sourceforge.net/
>
>
> Now I can double-click on the tutorial3.py from
> http://pyode.sourceforge.net/ and it simply runs!
>
>
> thanks,
>
> chris
>
>
>
Re: ODE sensitivity to displacement in space or time
Posted by chris <dr...@gmail.com>.
Glad to provide entertainment early in the new year :)
chris
On 1/4/07, Zubin Wadia <zw...@gmail.com> wrote:
> Hysterical! I thought he had some elaborate process flow for the experiment
> that he was invoking ODE for... oh well.
>
> Z.
>
> On 1/4/07, chris <dr...@gmail.com> wrote:
> >
> > Blast - sorry guys, wrong list (www.ode.org was not responding so I
> > thought it had moved, and the pyode ppl can't help)
> >
> > On 1/4/07, Holger Hoffstätte <ho...@wizards.de> wrote:
> > > Chris,
> > >
> > > I'm afraid you posted to the completely wrong list. Apache Ode is about
> > a
> > > BPEL process engine in Java, not a physics simulation environment. :)
> > >
> > > Holger
> > >
> >
>
>
Re: ODE sensitivity to displacement in space or time
Posted by Zubin Wadia <zw...@gmail.com>.
Hysterical! I thought he had some elaborate process flow for the experiment
that he was invoking ODE for... oh well.
Z.
On 1/4/07, chris <dr...@gmail.com> wrote:
>
> Blast - sorry guys, wrong list (www.ode.org was not responding so I
> thought it had moved, and the pyode ppl can't help)
>
> On 1/4/07, Holger Hoffstätte <ho...@wizards.de> wrote:
> > Chris,
> >
> > I'm afraid you posted to the completely wrong list. Apache Ode is about
> a
> > BPEL process engine in Java, not a physics simulation environment. :)
> >
> > Holger
> >
>
Re: ODE sensitivity to displacement in space or time
Posted by Zubin Wadia <zw...@gmail.com>.
Hysterical! I thought he had some elaborate process flow for the experiment
that he was invoking ODE for... oh well.
Z.
On 1/4/07, chris <dr...@gmail.com> wrote:
>
> Blast - sorry guys, wrong list (www.ode.org was not responding so I
> thought it had moved, and the pyode ppl can't help)
>
> On 1/4/07, Holger Hoffstätte <ho...@wizards.de> wrote:
> > Chris,
> >
> > I'm afraid you posted to the completely wrong list. Apache Ode is about
> a
> > BPEL process engine in Java, not a physics simulation environment. :)
> >
> > Holger
> >
>
Re: ODE sensitivity to displacement in space or time
Posted by chris <dr...@gmail.com>.
Blast - sorry guys, wrong list (www.ode.org was not responding so I
thought it had moved, and the pyode ppl can't help)
On 1/4/07, Holger Hoffstätte <ho...@wizards.de> wrote:
> Chris,
>
> I'm afraid you posted to the completely wrong list. Apache Ode is about a
> BPEL process engine in Java, not a physics simulation environment. :)
>
> Holger
>
Re: ODE sensitivity to displacement in space or time
Posted by chris <dr...@gmail.com>.
Blast - sorry guys, wrong list (www.ode.org was not responding so I
thought it had moved, and the pyode ppl can't help)
On 1/4/07, Holger Hoffstätte <ho...@wizards.de> wrote:
> Chris,
>
> I'm afraid you posted to the completely wrong list. Apache Ode is about a
> BPEL process engine in Java, not a physics simulation environment. :)
>
> Holger
>