Closed robotcopper closed 2 months ago
Hello @robotcopper!
Thank you so much for your great description of the issue. And you are right: It's a bug.
Do you want propose a PR fixing it?
Hello @robotcopper!
Thank you so much for your great description of the issue. And you are right: It's a bug.
Do you want propose a PR fixing it?
Hello @mateusmenezes95 , no problem. I made one. I tested it in open and closed loops and no regression appeared.
Fixed in #9
Hello @mateusmenezes95 ,
I am trying to use your controller in addition of the topic_based_ros2_control hardware interface.
The thing is that on my real robot i do not need of state interface. Indeed I am controlling my robot in open loop. So as you can expect, i set the
open_loop
param toTrue
and as i saw it in theomnidirectional_controller.cpp
it indeed unset the use of the state estimation tool.However, in the
omnidirectional_controller.cpp
the controller keep looking for astate_interface
even if theopen_loop
isTrue
.Or we could expect that only command would be send in
open_loop
, so in other words, just needingcommand_interface
in thehardware_interface
and not also astate_interface
am i right?So is it a bug? Did I miss something?
Here is what i expect to implement as![image](https://github.com/mateusmenezes95/omnidirectional_controllers/assets/95591405/d063c947-88e9-4a1f-818b-a068d51d566d)
hardware_interface
:PS: Thank you again for this very useful work you have made