Ask Your Question

creative_cimmons's profile - activity

2017-04-20 13:42:57 -0500 received badge  Notable Question (source)
2017-04-02 16:24:36 -0500 received badge  Nice Answer (source)
2016-06-17 06:50:31 -0500 received badge  Famous Question (source)
2016-06-17 06:50:31 -0500 received badge  Notable Question (source)
2016-06-17 06:50:31 -0500 received badge  Popular Question (source)
2016-04-14 04:37:48 -0500 received badge  Necromancer (source)
2016-03-17 00:19:11 -0500 received badge  Famous Question (source)
2016-02-04 10:19:27 -0500 received badge  Famous Question (source)
2016-01-27 04:18:57 -0500 received badge  Famous Question (source)
2015-12-10 07:42:31 -0500 received badge  Popular Question (source)
2015-12-09 22:23:06 -0500 received badge  Editor (source)
2015-12-09 22:22:20 -0500 asked a question compulsorily sourcing multiple setup files in same shell.


I have 2 catkin workspaces cata_ws , catb_ws

in cata_ws I have a.launch of nodea which when evoked would call b.launch which is in package nodeb in catb_ws.

my problem is if I do a rospack find ( in .launch I would do find) of nodeb , I get resource not found error.

The error is understandable because I have overwritten workspaces and currently cata_ws setup configuration is at the top.

So how do I successfully encode find of nodeb. I am working on a large project hence I need to compulsorily have separate workspaces.

Things I have tried and did not work.

  1. set args=nodeb and value=pathtonodeb in a.launch . The problem is that in b.launch there are many includes to launches of other nodes in catb_ws
  2. tried source [in cata_ws] devel/ --extends pathtocatb_ws/devel/ . Nope did not work

Thanks for the response.

2015-12-09 20:33:30 -0500 asked a question setup.bash refering to package and states file not found


I am a Ubuntu 14.04 user using ROS Indigo. I am not sure what happened recently but when I do [In catkin workspace] source devel/setup.bash I get the following error message

bash: /home/dinu94/catkin_ws/src/rocon/src/rocon_app_platform/rocon_app_utilities/shells/env.bash: No such file or directory
bash: /home/dinu94/catkin_ws/src/rocon/src/rocon_tools/rocon_launch/shells/env.bash: No such file or directory
bash: /home/dinu94/catkin_ws/src/rocon/src/rocon_multimaster/rocon_test/shells/env.bash: No such file or directory
bash: /home/dinu94/catkin_ws/src/rocon/src/rocon_tools/rocon_python_comms/completion/rospairbash: No such file or directory

I had installed rocon months before and deleted in quite recently. I don't know how this polluted setup.bash . I can't understand the code and hence unable to fix it. Any help would be appreciated .


2015-07-20 18:45:02 -0500 answered a question Tf has two or more unconnected trees

I have been working on it since May 15 and didn't find any such errors. Anyway if you did encounter errors write a simple tf publisher subscriber model where publisher is base_link and subscriber is map .

2015-07-19 21:29:02 -0500 answered a question tf Broadcaster Tutorial error

If it is your own package, make sure that it is included in cmakelists.txt

find_package(catkin REQUIRED COMPONENTS roscpp tf)

2015-07-15 15:33:53 -0500 received badge  Notable Question (source)
2015-07-15 15:32:26 -0500 received badge  Scholar (source)
2015-07-14 04:55:44 -0500 received badge  Notable Question (source)
2015-07-11 07:50:48 -0500 received badge  Nice Answer (source)
2015-07-09 20:53:11 -0500 received badge  Popular Question (source)
2015-07-08 22:02:39 -0500 asked a question cloud robotics engine for ROS


May I know if there exist a cloud platform for ROS computations. I know one 'Rapyuta' which was pretty much active 2 years ago. Now they have neither support for 14.04 nor for Indigo.


2015-07-08 18:52:46 -0500 commented question tf error: could not find connection between world and odom...

ok i guess I found it, it was simple assignment of param values. param name="odom_frame" value="odom"/ param name="world_frame" value="odom"/

2015-07-08 18:26:45 -0500 commented question tf error: could not find connection between world and odom...

mig, as i wont't have any drift errors ( UAV ) i thought of mapping map frame to odom frame instead of localization. ( unfortunately amcl only supports if its sensor is LIDAR based). Is there any easy route or do I have to write a listener-publisher of frames.

2015-07-01 17:19:00 -0500 commented answer Navigation Stack, AMCL or EKF

@paulbovel, what to do in the case I only use visual odometry and need amcl to output odom->map/world

2015-06-29 16:37:01 -0500 received badge  Popular Question (source)
2015-06-28 23:51:21 -0500 asked a question tf error: could not find connection between world and odom...


I am trying to make hector quadrotor use viso2_ros ,, after having configured correctly ( mono_odometer of viso2_ros takes in image from front_cam of hector) , I got warning that viso2_ros couldn't find the mapping base_link to front_cam_optical_frame and in the rviz of hector simulation, when I tried to add show pose parameter , it outputted 'couldn't find connection between world and odom...'

ok now I checked using tf view_frames

and this is what I got

image description

ok maybe it was because of the lack of link between base_link and fornt_cam_optical_frame,,

I linked both using some code

#include <ros/ros.h>
  #include <tf/transform_broadcaster.h>

  int main(int argc, char** argv){
 ros::init(argc, argv, "hector_remapping_tf");
 ros::NodeHandle n;

ros::Rate r(100);

 tf::TransformBroadcaster broadcaster;

       tf::Transform(tf::Quaternion(0, 0, 0, 1), tf::Vector3(0.0, 0.0, -0.2)),
       ros::Time::now(),"base_link", "front_cam_optical_frame"));

now the warning of lack of base_link to front_cam_optical_frame was gone but still the world and odom frame are not linked (checked using tf view_frames and rviz)

2015-06-26 18:13:17 -0500 commented question viso2_ros compilation error

any updates?

2015-06-24 02:55:08 -0500 received badge  Notable Question (source)
2015-06-22 02:19:57 -0500 received badge  Popular Question (source)
2015-06-20 22:48:02 -0500 asked a question converting one topic to another (front_cam/camera/image to sensor_msgs/image)


I am thinking of incorporating visual odometry onto hector-quadrotor which currently uses LIDAR. There is a front cam functionality ( topics like:


while the visual odometry (viso2_ros subscribes to sensor_msgs/image).

Is it possible to covert from front_cam/camera/image to sensor_msgs/image . Tried to look up for /front_cam but couldn't find any documentation relating to that


2015-06-16 15:32:11 -0500 received badge  Teacher (source)
2015-06-16 15:32:11 -0500 received badge  Necromancer (source)
2015-06-16 15:05:44 -0500 answered a question hector_quadrotor not listening to teleop_twist_keyboard

And here is the link to cpp version.

I am not sure about the 1000 m initialization,, I don't think that is required

2015-06-09 22:40:09 -0500 commented question hector_quadrotor not listening to teleop_twist_keyboard

I guess the problem is because of missing imu plugin,, could anybody debug that!!!

2015-06-09 15:39:07 -0500 commented question hector quadrotor controller

Install it as source

2015-06-09 15:15:54 -0500 commented question Hector quadrotor indoor tutorial simulation wont start

I am new to ros and ubuntu,,, I installed it on 8/6 ( building from source) and it didn't give any error. Sorry for lack of technical clarification but this worked for me.

2015-06-06 17:09:19 -0500 commented question hector_quadrotor not listening to teleop_twist_keyboard

Yeah, even I am facing the same issue!!!

2015-06-06 00:51:30 -0500 asked a question turning off hardware acceleration for rviz/hector quadrotor help


I am using ROS indigo on ubuntu 14.04 on virtualbox. Having set optimal settings in virtualbox (windows 8.1 host) ( enable 3d acceleration, using vguests addition ...) I have what I believe near optimal performance for the guest os. But when it comes to rviz

when I do rosrun rviz rviz

It spits out few errors such as vboxvideo not found and it outputs segementation fault core dumped and rviz exits.

Now the workaround given was it do export LIBGL_ALWAYS_SOFTWARE=1.

It is great for rviz but when running intensive applications which uses rviz like hector quadrotor,, this advice falls flat. The quadrotor is simply too slow.

If I didn't do the workaround, for rviz i get a segementation fault and for hector I get virtualbox referenced wrong memory address,, which is basically the same..

Is there any way where I could keep hardware acceleration and run rviz OR basically run hector on any other channel(you name it , I follow it)


2015-06-04 16:25:43 -0500 received badge  Enthusiast
2015-06-01 14:39:00 -0500 received badge  Supporter (source)