Thread: [Pyode-user] PATCH: LMotor

SourceForge Headquarters 225 Broadway Suite 1600 San Diego, CA 92101 +1 (858) 422-6466

Adds support for the LMotor motor, which is in ODE 0.7. This motor
parallels the AMotor code, so I copied and search-replaced. Please apply.

Ethan

Yikes, looks like I goofed. The getMode/setMode functions should be
removed; they don't have any equivalent for LMotors. Also, the 'rel'
argument is currently ignored in the LMotor class; a note should be
made of that in the docstring.

A new patch is attached that addresses these issues.

Sorry about that!

Ethan

i think the FixedJoint is missing setParameter for doing custom cfm and erp.  also i've noticed
that at least for some joint types, getfeedback works, but setfeedback does not.  anybody else
having the same problems?
-brett

____________________________________________________________________________________
Cheap Talk? Check out Yahoo! Messenger's low PC-to-Phone call rates 
(http://voice.yahoo.com)

Hart's Antler wrote:
> i think the FixedJoint is missing setParameter for doing custom cfm and=
 erp.  also i've noticed
> that at least for some joint types, getfeedback works, but setfeedback =
does not.  anybody else
> having the same problems?
> -brett

Hi,

I don't see any dJointGetFixedParam/dJointSetFixedParam functions in
ODE 0.7. Likewise, I don't see any place for the parameters in the
dxFixedJoint structure. So, the functions are missing upstream.

Ethan

Ethan Glasser-Camp wrote:
> Yikes, looks like I goofed. The getMode/setMode functions should be

Jeez, can you tell I never used pyrex before?

This time the patch actually builds, and appears to work.

Ethan

Ethan Glasser-Camp wrote:
> Ethan Glasser-Camp wrote:
>> Yikes, looks like I goofed. The getMode/setMode functions should be
> 
> Jeez, can you tell I never used pyrex before?
> 
> This time the patch actually builds, and appears to work.

I've just applied and committed the patch. I've also applied the patch 
that Leonard Ritter has posted a while ago that adds support for the 
Plane2DJoint.

Finally, I've upgraded to ODE 0.7 as well (Windows only, using the 
binaries). I haven't committed the updated setup script yet. It seems 
that trimeshes are now always part of ODE and we don't have to 
distinguish between the two versions anymore. Is this correct? (this 
would simplify the setup script quite a bit).

Are there any other mails I've overlooked that also contain patches 
you'd like to be applied?

- Matthias -

Matthias Baas wrote:
> Ethan Glasser-Camp wrote:
>> Ethan Glasser-Camp wrote:
>>> Yikes, looks like I goofed. The getMode/setMode functions should be
>> Jeez, can you tell I never used pyrex before?
>>
>> This time the patch actually builds, and appears to work.
>=20
> I've just applied and committed the patch. I've also applied the patch =

> that Leonard Ritter has posted a while ago that adds support for the=20
> Plane2DJoint.

There is still at least one (minor) thing wrong with LMotor patch I
wrote, but I'll address these in a further patch.

> Are there any other mails I've overlooked that also contain patches=20
> you'd like to be applied?

I'd like to write support for Cylinder and change CCylinder to
Capsule, but I haven't written these patches yet.

Ethan

Ethan Glasser-Camp wrote:
> There is still at least one (minor) thing wrong with LMotor patch I
> wrote, but I'll address these in a further patch.

Here -- rel is actually not ignored in ODE 0.7, there's just a comment
in the source code that says it is. And I added LMotor to the list of
classes in ode.pyx.

Ethan

Ethan Glasser-Camp wrote:
> I'd like to write support for Cylinder and change CCylinder to
> Capsule, but I haven't written these patches yet.

The attached patch should change CCylinder to Capsule, but keep the
GeomCCylinder name around for backwards compatibility.

Ethan

Ethan Glasser-Camp wrote:
> -    dGeomID dCreateCCylinder (dSpaceID space, dReal radius, dReal length)
> +    dGeomID dCreateCapsule (dSpaceID space, dReal radius, dReal length)

Is there actually a difference between those functions (except the 
name)? Or are those just aliases for the same functions. Will the 
ccylinder functions ever be removed from the API?

- Matthias -

[Sorry Matthias for the triplicate post! Yeesh, where is my brain?]

Matthias Baas wrote:
> Finally, I've upgraded to ODE 0.7 as well (Windows only, using the=20
> binaries). I haven't committed the updated setup script yet. It seems=20
> that trimeshes are now always part of ODE and we don't have to=20
> distinguish between the two versions anymore. Is this correct? (this=20
> would simplify the setup script quite a bit).

Hi,

First, thanks for applying my patches. :)

Second, it looks to me like you can still run ODE's ./configure with a
--disable-opcode, which disables trimesh support. But I admit I don't
really know a lot about either codebase. I, too, would prefer if
trimesh support was mandatory, but I'm not sure if this is actually
the case.

Ethan

Ethan Glasser-Camp wrote:
> [Sorry Matthias for the triplicate post! Yeesh, where is my brain?]

No problem. :)

> Matthias Baas wrote:
>> Finally, I've upgraded to ODE 0.7 as well (Windows only, using the 
>> binaries). I haven't committed the updated setup script yet. It seems 
>> that trimeshes are now always part of ODE and we don't have to 
>> distinguish between the two versions anymore. Is this correct? (this 
>> would simplify the setup script quite a bit).
> 
> Hi,
> 
> First, thanks for applying my patches. :)

You're welcome.

> Second, it looks to me like you can still run ODE's ./configure with a
> --disable-opcode, which disables trimesh support. But I admit I don't
> really know a lot about either codebase. I, too, would prefer if
> trimesh support was mandatory, but I'm not sure if this is actually
> the case.

I suppose if we'd also want to support ODE without trimesh support we 
would have to parse ode/config.h and check the dTRIMESH_ENABLED macro.

- Matthias -

Thread: [Pyode-user] PATCH: LMotor

pyode-user