Hallo,
I wrote a viewer program for kinematic chains using KDL's Python bindings.
You can get it from this webpage:
http://www9.in.tum.de/~kleinehe/roboview.html
It's based on VTK
bindings of these two libraries, as well as KDL's Python bindings.
I tested it against KDL SVN, but there's a small bug in the Python bindings for
which I attached a patch. I made another patch that fixes some complilation
problem I'm having on my machine.
If you're interested please test the program. I'm always happy about feedback.
-- Jonathan
Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm
New Viewer for KDL Chains written in Python
On Thu, 4 Sep 2008, Jonathan Kleinehellefort wrote:
> I wrote a viewer program for kinematic chains using KDL's Python bindings.
> You can get it from this webpage:
> http://www9.in.tum.de/~kleinehe/roboview.html
> It's based on VTK
> bindings of these two libraries, as well as KDL's Python bindings.
>
> I tested it against KDL SVN, but there's a small bug in the Python
> bindings for which I attached a patch. I made another patch that fixes
> some complilation problem I'm having on my machine.
>
> If you're interested please test the program. I'm always happy about feedback.
>
Your webpage doesn't tell me anything about the (long term) purpose of your
project. So, I ask myself the question why you want to start all over again
from scratch, instead of using _much_ more developed projects like Blender,
that have already built-in support for kinematic chains?
And VTK is first of all targeting static, large-scale data visualisation,
so not the best infrastructure project to build upon.
With all respect, but I have the strong feeling that our project will never
be able to reach a critical mass... Of course, feel free to prove me
wrong:-)
Best regards,
Herman
New Viewer for KDL Chains written in Python
On Thu, Sep 04, 2008 at 04:28:22PM +0200, Herman Bruyninckx wrote:
> Your webpage doesn't tell me anything about the (long term) purpose of your
> project. So, I ask myself the question why you want to start all over again
> from scratch, instead of using _much_ more developed projects like Blender,
> that have already built-in support for kinematic chains?
Yes, I should add something about its intended use there.
It's just for seeing (and testing) chain descriptions done with KDL. I just
did it to visualize a KDL description I had and then extended it so it could
be used by another person.
I'm not really planing on adding a lot of features to it at the moment, I will
just fix bugs that might come up.
I will of course consider suggestions for improvements.
> And VTK is first of all targeting static, large-scale data visualisation,
> so not the best infrastructure project to build upon.
This is true, VTK is not really helping here, but I first did a small
visualization using it and then extended it a bit. Also I wanted to learn
using it.
> With all respect, but I have the strong feeling that our project will never
> be able to reach a critical mass... Of course, feel free to prove me
> wrong:-)
I'm not really trying to get many users. I thought since I have the code
anyway, I might as well publish it.
-- Jonathan
New Viewer for KDL Chains written in Python
On Thu, 4 Sep 2008, Jonathan Kleinehellefort wrote:
> On Thu, Sep 04, 2008 at 04:28:22PM +0200, Herman Bruyninckx wrote:
>> Your webpage doesn't tell me anything about the (long term) purpose of your
>> project. So, I ask myself the question why you want to start all over again
>> from scratch, instead of using _much_ more developed projects like Blender,
>> that have already built-in support for kinematic chains?
>
> Yes, I should add something about its intended use there.
>
> It's just for seeing (and testing) chain descriptions done with KDL. I just
> did it to visualize a KDL description I had and then extended it so it could
> be used by another person.
>
> I'm not really planing on adding a lot of features to it at the moment, I will
> just fix bugs that might come up.
>
> I will of course consider suggestions for improvements.
>
>> And VTK is first of all targeting static, large-scale data visualisation,
>> so not the best infrastructure project to build upon.
>
> This is true, VTK is not really helping here, but I first did a small
> visualization using it and then extended it a bit. Also I wanted to learn
> using it.
>
Nevertheless, Blender would be a much better choice... You can use Python
scripts in Blender to add your own code rather easily.
>> With all respect, but I have the strong feeling that our project will never
>> be able to reach a critical mass... Of course, feel free to prove me
>> wrong:-)
>
> I'm not really trying to get many users. I thought since I have the code
> anyway, I might as well publish it.
That's always good! :-) Thanks.
Herman