pyode-user — PyODE User Discussions

[Pyode-user] os x fat binary build

[Diez B. R. <de...@we...>]

Hi,

I had troubles getting PyODE to work under OSX tiger with my intel-mac 
because the ODE-lib wasn't compiling as fat-binary. So I create a osx 
framework build with XCode, and build PyODE against. ODE version is 0.7

You can get things from here:

http://www.roggisch.de/PyODE_osx_framework_fatbinary.tar.bz2

So far, no manuals/install-instructions whatsoever. You just need to add 
the ODE.framework to some framework-folder, like ~/Library/Frameworks, 
and easy_install the egg.

Maybe this is of use for someone.

Diez

Re: [Pyode-user] Pixel-perfect, again

[Chris B. <chr...@gm...>]

On 21/02/07, Andr=E1s =C1cs <a....@in...> wrote:
> I am making a 2D gravity shooter, and not
> sure how to integrate pixel-perfect collision
> (a terrain) into a simulation.
> I read that a collider could be plugged into ode.
> How can this be done?

I assume you have a collision detector that detects ODE geoms
penetrating your terrain? If so, all you need to do is to call this
collider to generate contacts, then convert them into ode.Contact
instances, then for each of those create an ode.ContactJoint and
attach it to the penetrating Geom. But that's relatively easy, the
hard part is writing the collision detector. If you don't have one,
you'll need to figure out how to detect contacts and contact direction
etc. between whatever terrain model you're using and each ODE Geom.

Re: [Pyode-user] get angles of universal joints?

[Chris B. <chr...@gm...>]

On 21/02/07, Ethan Glasser-Camp <gl...@cs...> wrote:
> Thomas R=FCckstie=DF wrote:
> > dJointGetUniversalAngle1()
> > dJointGetUniversalAngle2()
> >
> > but I couldn't find a PyODE aequivalent wrapper function anywhere.
> > Help would be appreciated! Thanks.

You can also create an AMotor, attach it to the same bodies as the
universal joint, and then call
AMotor.getAngle(x). The motor will be inactive unless you ask it to do
something.

[Pyode-user] Small bug in Tutorial 3

[<Tho...@di...>]

Hi,

the pyode tuts are a charming way to make first contact, but at tutorial 
three, I wasn't sure exactly _what_ I was seeing, until I replaced
glutInitDisplayMode (GLUT_RGB | GLUT_DOUBLE)
with
glutInitDisplayMode (GLUT_RGB | GLUT_DOUBLE | GLUT_DEPTH)
for the obvious reasons :-).

Regards,
    Thomas

Re: [Pyode-user] get angles of universal joints?

[Ethan Glasser-C. <gl...@cs...>]

Thomas Rückstieß wrote:
> dJointGetUniversalAngle1()
> dJointGetUniversalAngle2()
> 
> but I couldn't find a PyODE aequivalent wrapper function anywhere. 
> Help would be appreciated! Thanks.

Hi,

I didn't see these functions provided in Pyode, but here's an untested
patch which should provide those functions, as well as the
Angle{1,2}Rate functions, as well as fix the indentation in
UniversalJoint.getParam.

Hope this helps,

Ethan

[Pyode-user] Pixel-perfect, again

[<a....@in...>]

Hi=21

I am making a 2D gravity shooter, and not
sure how to integrate pixel-perfect collision
(a terrain) into a simulation.
I read that a collider could be plugged into ode.
How can this be done?

The terrain images the game generates are quite large
(max. 2048*2048), so I welcome optimization ideas too:)

Thanks very-very much in advance for any help.

Andras

[Pyode-user] get angles of universal joints?

[<rue...@in...>]

Hi everybody,

does anyone know if it is possible to get the angles of a universal =20
joint under PyODE? The original ODE functions are

dJointGetUniversalAngle1()
dJointGetUniversalAngle2()

but I couldn't find a PyODE aequivalent wrapper function anywhere.
Help would be appreciated! Thanks.

Thomas

--
Dipl.-Inf. Thomas R=FCckstie=DF
Technische Universit=E4t M=FCnchen
Lehrstuhl Informatik VI - Robotik und Echtzeitsysteme

Raum 00.09.064
Tel. (089) 289 17850
rue...@in...

[Pyode-user] [PATCH] __delattr__ for geoms

[Ethan Glasser-C. <gl...@cs...>]

This patch allows statements like "del geom.attr".

I wasn't sure whether or not to add a trailing period in the
AttributeError message. The Python AttributeError message doesn't have
it, so it looks weird -- on the other hand, it also adds a distinct
signifier for pyode error messages. I decided to stick with the period.

Ethan

[Pyode-user] [PATCH] Docstrings for GeomRay

[Ethan Glasser-C. <gl...@cs...>]

Hi guys,

Thanks for the 1.2.0 release the other day! It really made my day.

Here's a patch that adds some docstrings to the GeomRay class. They
aren't much but they might help someone looking for those functions.

Ethan

Re: [Pyode-user] API site down

[Timothy S. <tim...@gm...>]

Hi Jeff,

Since the new PyODE 1.2.0 release, the API reference is now at
http://pyode.sourceforge.net/api-1.2.0/index.html for 1.2.0 and at
http://pyode.sourceforge.net/api-1.1.0/index.html for 1.1.0.

--
Timothy

[Pyode-user] API site down

[Jeff B. <its...@gm...>]

The API for pyode seems to be down.  When I try to access it, I get this:

> 1. *Server:* pyode.sourceforge.net
> 2. *URL path:* /api/index.html
> 3. *Error notes:* File does not exist:
> /home/groups/p/py/pyode/htdocs/api/index.html
> 4. *Error type:* 404
> 5. *Request method:* GET
> 6. *Request query string:*
> 7. *Time:* 2007-02-12 11:05:01 PST (1171307101)


Is there anywhere else I can get an API reference?

Re: [Pyode-user] building with gcc 4.1

[Timothy S. <tim...@gm...>]

On 2/9/07, Dmitry Marakasov <am...@am...> wrote:
> I found out that pyrex 0.9.4.1 (as opposed to 0.9.3) does generate
> correct C code, so I suggest py-ode developers to regenerate *.c
> in the distribution with newer pyrex and release a new tarball.

We released PyODE 1.2.0 a few days ago. It was built with Pyrex
0.9.5.1a so it should work with GCC 4.1.

--
Timothy

Re: [Pyode-user] building with gcc 4.1

[Dmitry M. <am...@am...>]

* Diez B. Roggisch (die...@ar...) wrote:
> On Wednesday 31 January 2007 12:23, Dmitry Marakasov wrote:
> > I'm maintaining FreeBSD port of py-ode. There are no problems when
> > building it with gcc 3.4, but when I try to use gcc 4.1 (which will be
> > introduced in new FreeBSD versions) instead, I get the following errors:
> >
> > Any clues on how this can be fixed?
> 
> Just a guess: You might be better off asking that on the pyrex mailing lists, 
> as PyODE uses pyrex to generate the C-code.

Thanks for the hint!
I found out that pyrex 0.9.4.1 (as opposed to 0.9.3) does generate
correct C code, so I suggest py-ode developers to regenerate *.c
in the distribution with newer pyrex and release a new tarball.

As a temporary fix, one can just delete *.c files after unpacking
py-ode, so they'll be regenerated when running setup.py.
Pyrex >= 0.9.4.1 needs to be installed for that, of course.

Thanks again for help.

-- 
Best regards,
  Dmitry Marakasov               mailto:am...@am...

[Pyode-user] precision used in pyode

[chris <dra...@gm...>]

Hi, I have installed PyODE-1.1.0.win32-py2.4.exe
 from http://pyode.sourceforge.net/
and want to know to what precision the ode was compiled for: single or
double floats,

thanks,

chris

Re: [Pyode-user] building with gcc 4.1

[Diez B. R. <die...@ar...>]

On Wednesday 31 January 2007 12:23, Dmitry Marakasov wrote:
> Hi!
>
> I'm maintaining FreeBSD port of py-ode. There are no problems when
> building it with gcc 3.4, but when I try to use gcc 4.1 (which will be
> introduced in new FreeBSD versions) instead, I get the following errors:
>
> Any clues on how this can be fixed?

Just a guess: You might be better off asking that on the pyrex mailing lists, 
as PyODE uses pyrex to generate the C-code.

Diez

[Pyode-user] building with gcc 4.1

[Dmitry M. <am...@am...>]

Hi!

I'm maintaining FreeBSD port of py-ode. There are no problems when
building it with gcc 3.4, but when I try to use gcc 4.1 (which will be
introduced in new FreeBSD versions) instead, I get the following errors:

Any clues on how this can be fixed?

--- cut ---
ode_trimesh.c is up to date
ode_notrimesh.c is up to date
Installing without trimesh support.
running build
running build_py
running build_ext
building 'ode' extension
gcc41 -fno-strict-aliasing -DNDEBUG -O2 -fno-strict-aliasing -pipe -march=pentium4 -D__wchar_t=wchar_t -DTHREAD_STACK_SIZE=0x20000 -O2 -fno-strict-aliasing -pipe -march=pentium4 -fPIC -I/usr/X11R6/include -I/usr/local/include/python2.4 -c ode_notrimesh.c -o build/temp.freebsd-6.1-RELEASE-p12-i386-2.4/ode_notrimesh.o
ode_notrimesh.c: In function '__pyx_f_3ode_4Body_getMass':
ode_notrimesh.c:3776: error: invalid lvalue in assignment
ode_notrimesh.c:3784: error: invalid lvalue in assignment
ode_notrimesh.c: In function '__pyx_f_3ode_9BallJoint___new__':
ode_notrimesh.c:5973: error: invalid lvalue in assignment
ode_notrimesh.c:5988: error: invalid lvalue in assignment
ode_notrimesh.c: In function '__pyx_f_3ode_10HingeJoint___new__':
ode_notrimesh.c:6272: error: invalid lvalue in assignment
ode_notrimesh.c:6287: error: invalid lvalue in assignment
ode_notrimesh.c: In function '__pyx_f_3ode_11SliderJoint___new__':
ode_notrimesh.c:6747: error: invalid lvalue in assignment
ode_notrimesh.c:6762: error: invalid lvalue in assignment
ode_notrimesh.c: In function '__pyx_f_3ode_14UniversalJoint___new__':
ode_notrimesh.c:7084: error: invalid lvalue in assignment
ode_notrimesh.c:7099: error: invalid lvalue in assignment
ode_notrimesh.c: In function '__pyx_f_3ode_11Hinge2Joint___new__':
ode_notrimesh.c:7601: error: invalid lvalue in assignment
ode_notrimesh.c:7616: error: invalid lvalue in assignment
ode_notrimesh.c: In function '__pyx_f_3ode_10FixedJoint___new__':
ode_notrimesh.c:8196: error: invalid lvalue in assignment
ode_notrimesh.c:8211: error: invalid lvalue in assignment
ode_notrimesh.c: In function '__pyx_f_3ode_12ContactJoint___new__':
ode_notrimesh.c:8326: error: invalid lvalue in assignment
ode_notrimesh.c:8342: error: invalid lvalue in assignment
ode_notrimesh.c: In function '__pyx_f_3ode_6AMotor___new__':
ode_notrimesh.c:8439: error: invalid lvalue in assignment
ode_notrimesh.c:8454: error: invalid lvalue in assignment
ode_notrimesh.c: In function '__pyx_f_3ode_11SimpleSpace___new__':
ode_notrimesh.c:10682: error: invalid lvalue in assignment
ode_notrimesh.c:10696: error: invalid lvalue in assignment
ode_notrimesh.c: In function '__pyx_f_3ode_9HashSpace___new__':
ode_notrimesh.c:10771: error: invalid lvalue in assignment
ode_notrimesh.c:10785: error: invalid lvalue in assignment
ode_notrimesh.c: In function '__pyx_f_3ode_13QuadTreeSpace___new__':
ode_notrimesh.c:10969: error: invalid lvalue in assignment
ode_notrimesh.c:10983: error: invalid lvalue in assignment
ode_notrimesh.c: In function '__pyx_f_3ode_10GeomSphere___new__':
ode_notrimesh.c:11195: error: invalid lvalue in assignment
ode_notrimesh.c:11209: error: invalid lvalue in assignment
ode_notrimesh.c: In function '__pyx_f_3ode_7GeomBox___new__':
ode_notrimesh.c:11455: error: invalid lvalue in assignment
ode_notrimesh.c:11469: error: invalid lvalue in assignment
ode_notrimesh.c: In function '__pyx_f_3ode_9GeomPlane___new__':
ode_notrimesh.c:11768: error: invalid lvalue in assignment
ode_notrimesh.c:11782: error: invalid lvalue in assignment
ode_notrimesh.c: In function '__pyx_f_3ode_13GeomCCylinder___new__':
ode_notrimesh.c:12066: error: invalid lvalue in assignment
ode_notrimesh.c:12080: error: invalid lvalue in assignment
ode_notrimesh.c: In function '__pyx_f_3ode_7GeomRay___new__':
ode_notrimesh.c:12346: error: invalid lvalue in assignment
ode_notrimesh.c:12360: error: invalid lvalue in assignment
ode_notrimesh.c: In function '__pyx_f_3ode_13GeomTransform___new__':
ode_notrimesh.c:12650: error: invalid lvalue in assignment
ode_notrimesh.c:12664: error: invalid lvalue in assignment
ode_notrimesh.c: In function '__pyx_f_3ode_collide':
ode_notrimesh.c:13091: error: invalid lvalue in assignment
ode_notrimesh.c:13138: error: invalid lvalue in assignment
error: command 'gcc41' failed with exit status 1
*** Error code 1
--- cut ---

-- 
Best regards,
  Dmitry Marakasov               mailto:am...@am...

Re: [Pyode-user] about ode 0.7

[Patrick M. <sal...@gm...>]

Hi, I would really like the python2.5 windows bindings as well, if
possible.  I am curious how long it will be before the next version of pyode
is officially released.  How much more is left to be done?

[Pyode-user] libode.dylib linking issue on macosx + darwinports

[Casey D. <cas...@gm...>]

I had a problem using pyode (from cvs) on MacOS X using python
2.4 and ode 0.7 installed via darwinports. The setup.py was able to
build and install pyode successfully, but I could not import ode:

>>> import ode
Traceback (most recent call last):
 File "<stdin>", line 1, in ?
ImportError: Failure linking new module:
/opt/local/lib/python2.4/site-packages/ode.so: Library not loaded:
libode.dylib  Referenced from:
/opt/local/lib/python2.4/site-packages/ode.so
 Reason: image not found

Looking at the ode.so in site-packages shows that libode.dylib was
linked in with a relative path:

% otool -L /opt/local/lib/python2.4/site-packages/ode.so
/opt/local/lib/python2.4/site-packages/ode.so:
       libode.dylib (compatibility version 0.0.0, current version 0.0.0)
       /usr/lib/libstdc++.6.dylib (compatibility version 7.0.0,
current version 7.4.0)
       /usr/lib/libgcc_s.1.dylib (compatibility version 1.0.0, current
version 1.0.0)
       /usr/lib/libSystem.B.dylib (compatibility version 1.0.0,
current version 88.3.4)

If I diddled PYLD_LIBRARY_PATH, I could get ode to import, but it
broke other things in the process. Ultimately I was able to fix it be
changing the link to libode.dylib to an absolute path.

% sudo install_name_tool -change libode.dylib
/opt/local/lib/libode.dylib
/opt/local/lib/python2.4/site-packages/ode.so

I'm not sure the best way to fixup setup.py et al to do this
automagically (i.e., just prevent the unqualified link in the first
place). If you do great, if not I can do some more digging and figure
it out.

Ideally pyode would have a darwinport itself, though you'll need to
release what's in cvs first for that to be practical ;^)

Out of curiousity what is the state of the version in cvs? How close
to stable/release ready would you consider it?

Thanks much for this package, I'm a newbie to ode but it looks like a
great tool.

-Casey

Re: [Pyode-user] about ode 0.7

[Timothy S. <tim...@gm...>]

On 1/10/07, Matthias Baas <ba...@ir...> wrote:
> Timothy, have you some time left for a new release?

I'm away from my Linux computer at the moment and I'll only be getting
back in around 2.5 weeks. I'm sorry but I won't be able to help with a
release until then.

--
Timothy

Re: [Pyode-user] about ode 0.7

[Jacob E. <jac...@gm...>]

You might consider just using cmake. (Cross-Platform Make)
http://www.cmake.org/HTML/Index.html

This will generate build scripts for your system, find all the
dependencies, and work on several different operating systems.  It's
supposed to relieve you of many of the interoperatiblity issues you
have to think about.

One of my coworkers added cmake as the build environment to our
project, and we successfully build it on windows, OSX, and Linux, each
of us developing on our platform of choice.

Jacob Everist

On 1/11/07, Matthias Baas <ba...@ir...> wrote:
> Ethan Glasser-Camp wrote:
> > If you could even send me Windows binaries for pyode CVS I'd be most
> > grateful.
>
> No problem. Are you using Python 2.4 or 2.5?
>
> > What do you mean "default paths"? You mean where things are installed
> > after build?
>
> No (that's determined by the distutils). I mean the paths where the
> compiler will look for the ODE headers and lib.
> Currently, the setup script just distinguishes between Windows and
> non-Windows and on non-Windows system a couple of default locations are
> used. The question was just if I should also distinguish between Linux
> and OSX, but I don't think this is necessary here.
>
> - Matthias -
>
>
> -------------------------------------------------------------------------
> Take Surveys. Earn Cash. Influence the Future of IT
> Join SourceForge.net's Techsay panel and you'll get the chance to share your
> opinions on IT & business topics through brief surveys - and earn cash
> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
> _______________________________________________
> Pyode-user mailing list
> Pyo...@li...
> https://lists.sourceforge.net/lists/listinfo/pyode-user
>


-- 

Jacob Everist

Re: [Pyode-user] about ode 0.7

[Matthias B. <ba...@ir...>]

Ethan Glasser-Camp wrote:
> If you could even send me Windows binaries for pyode CVS I'd be most
> grateful. 

No problem. Are you using Python 2.4 or 2.5?

> What do you mean "default paths"? You mean where things are installed
> after build? 

No (that's determined by the distutils). I mean the paths where the 
compiler will look for the ODE headers and lib.
Currently, the setup script just distinguishes between Windows and 
non-Windows and on non-Windows system a couple of default locations are 
used. The question was just if I should also distinguish between Linux 
and OSX, but I don't think this is necessary here.

- Matthias -

Re: [Pyode-user] about ode 0.7

[Matthias B. <ba...@ir...>]

Jacob Everist wrote:
> Usually, /opt/local will work okay for linux.  /opt is not a commonly
> used path, but some packages insist on using it, and so linux can
> handle it just fine.
> 
> Is /opt what's used for OSX?  I'm just not sure where this path comes
> from since I'm just a linux guy.

/opt/local is the default path where the MacPorts software (formerly 
DarwinPorts) is installed. This is just what I used to install ODE but 
it's not necessarily what other OSX users will use as well (you could 
also use Fink or compile ODE yourself in which case ODE might get 
installed in /sw or /usr/local).

- Matthias -

Re: [Pyode-user] about ode 0.7

[Ethan Glasser-C. <gl...@cs...>]

Matthias Baas wrote:
> It compiles here on Windows and OSX (is it ok when I add /opt/local to =

> the default paths on non-Windows builds or would this break something=20
> for Linux people?), so I can provide Windows binaries. Timothy, have yo=
u=20
> some time left for a new release?

If you could even send me Windows binaries for pyode CVS I'd be most
grateful. My only Windows machine right now is an old Win98 clunker
and I don't have any Microsoft compiler tools, so I'm left trying to
build using mingw32 and dev-cpp, and frankly I'm not up to the task.

What do you mean "default paths"? You mean where things are installed
after build? My system (Debian unstable) doesn't have anything under
/opt, especially not /opt/local. I'm willing to test packages on my
system, though.

Ethan

Re: [Pyode-user] about ode 0.7

[Jacob E. <jac...@gm...>]

Usually, /opt/local will work okay for linux.  /opt is not a commonly
used path, but some packages insist on using it, and so linux can
handle it just fine.

Is /opt what's used for OSX?  I'm just not sure where this path comes
from since I'm just a linux guy.

Jacob Everist

On 1/10/07, Matthias Baas <ba...@ir...> wrote:
> Ethan Glasser-Camp wrote:
> > Is a new release of pyode expected any time soon? Is there anything
> > holding it back from being released?
>
> Nothing particular, except that I and Timothy have just been busy with
> other things (and nobody has requested a new release yet)... ;)
>
> It compiles here on Windows and OSX (is it ok when I add /opt/local to
> the default paths on non-Windows builds or would this break something
> for Linux people?), so I can provide Windows binaries. Timothy, have you
> some time left for a new release?
>
> - Matthias -
>
> -------------------------------------------------------------------------
> Take Surveys. Earn Cash. Influence the Future of IT
> Join SourceForge.net's Techsay panel and you'll get the chance to share your
> opinions on IT & business topics through brief surveys - and earn cash
> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
> _______________________________________________
> Pyode-user mailing list
> Pyo...@li...
> https://lists.sourceforge.net/lists/listinfo/pyode-user
>


-- 

Jacob Everist

Re: [Pyode-user] about ode 0.7

[Matthias B. <ba...@ir...>]

Ethan Glasser-Camp wrote:
> Is a new release of pyode expected any time soon? Is there anything
> holding it back from being released? 

Nothing particular, except that I and Timothy have just been busy with 
other things (and nobody has requested a new release yet)... ;)

It compiles here on Windows and OSX (is it ok when I add /opt/local to 
the default paths on non-Windows builds or would this break something 
for Linux people?), so I can provide Windows binaries. Timothy, have you 
some time left for a new release?

- Matthias -