Just as a reference, here's an example of how the yaml file currently is structured:
Plugins:
- type: "image"
class_name: "rviz/Image"
rviz_name: "Temoto Image"
data_type: "sensor_msgs/Image"
- type: "laser_scan"
class_name: "rviz/LaserScan"
rviz_name: "Temoto Laser Scan"
data_type: "sensor_msgs/LaserScan"
So the first problem is that the field names in the yaml description are quite confusing, so i propose:
- we should get rid of the
type
class_name -> plugin_name
rviz_name -> name_seen_by_user
data_type -> visualized_data
Which brings us to the second issue, lack of support for multiple topics, thus I think the visualized_data should be a vector of key value pairs (example):
visualized_data:
- laserscan: "sensor_msgs/LaserScan"
or
visualized_data:
- panospheric_image_front: "sensor_msgs/Image"
- panospheric_image_back: "sensor_msgs/Image"
This would require some changes both in the visualization_manager.cpp and visualization_manager_interface.h so that user can specify the topics.
Just as a reference, here's an example of how the yaml file currently is structured:
So the first problem is that the field names in the yaml description are quite confusing, so i propose:
typeclass_name->plugin_namerviz_name->name_seen_by_userdata_type->visualized_dataWhich brings us to the second issue, lack of support for multiple topics, thus I think the
visualized_datashould be a vector of key value pairs (example):or
This would require some changes both in the
visualization_manager.cppandvisualization_manager_interface.hso that user can specify the topics.