kolya_rage's profile - activity

Hi, I'm trying to run move_group code through gdb. I compiled moveit from sources with --cmake-args -DCMAKE_BUILD_TYPE=Debug. Still, it seems like debug information haven't been lost. What I did: I ran my own node that uses move_group.plan(my_plan) and tried to step through plan(). But it doesn't step into the function like it hadn't written debugging data. I use ros kinetic Would use any help. Thanks

tell please. Is this actual tutorial's code you are testing, or your own implementation. If it is - please mail me with

So you are saying that you know dimensions and locations of your obstacle, and now you want your robot to know those to

It has two cameras. So you can get depth from a single frame

orb slam, lsd slam

Yes, you can.

well, I suggest either updating to ubuntu 16 or working through docker which is the best thing ever if you have differen

well, I suggest either updating to ubuntu 16 or working through docker which is the best thing ever if you have differen

I tried docker container with ubuntu 14.04. Installation went just fine. apt update apt install libgl1-mesa-dev-lts-tr

Yes, this is a problem which is hard to debug, because function MoveGroup::plan() returns either success or fail(besides

Yes, this is a problem which is hard to debug, because function MoveGroup::plan() returns either success or fail(besides

Yes, this is a problem which is hard to debug, because function MoveGroup::plan() returns either success or fail(besides

Yes, this is a problem which is hard to debug, because function MoveGroup::plan() returns either success or fail(besides

Yes, this is a problem which is hard to debug, because function MoveGroup::plan() returns either success or fail(besides

Yes, this is a problem which is hard to debug, because function MoveGroup::plan() returns either success or fail(besides

Yes, this is a problem which is hard to debug, because function MoveGroup::plan() returns either success or fail(besides

Yes, this is a problem which is hard to debug, because function MoveGroup::plan() return either success or fail(besides

this is weird. The package is listed at ubuntu website. You sure you did sudo apt-update? I can try running in docker i

If you can show a video capture of rviz - would help a lot

Yes, you can. You can either plan for a specific joint state(link to plan struct) as a start state. Or if the trajectory

simpy making a vector of poses already would be a graph. You look to nicely visualize your graph, or?

It should. So you saved your joint_state before execution, then published a new one and then executed once again?

You already have made an execution, It means the robot has already moved, and applying the same planning won't work, bec

You already have made an execution, It means the robot has already moved, and applying the same planning won't work, bec

So yes. I you already have made an execution, It means the robot has already moved, and applying the same planning won't

well, every thing seems right to me. Does planning execution works in the rviz? I mean if you do planning in rviz moveit

I see that wasn't a good advice for me. so fake execution is actually robot simulation.

try setting fake execution to false

I will try myself the tutorial a bit later

try to set fake execution to false

I see, you are following a tutorial. Well, the code above is ok, there are no mistakes. Show me your launch file to run

I see, you are following a tutorial. Well, the code above is ok, there are no mistakes. Show me you launch file to run t

you didn't execute you trajectory at the first time, you only planned. but then yes - trajectory was executed. Are you u

https://github.com/IntelRealSense/librealsense/wiki/Projection-in-RealSense-SDK-2.0 Here is projection and deprojection

Your point cloud is produced by the real sense, right?

I see. So you should have projection matrix from the realsense camera, don't you?

I'm not sure I understand. You want to project rgb to pcd. Or you want to downscale point cloud somehow?

this is weird. try rospy.sleep(1.). Totally not sure if this will help :)

Check that you have a proper header: it should have time stamp, so the ROS thinks the data is fresh. Make sure you do so

you tried this one https://askubuntu.com/questions/389903/ibus-doesnt-seem-to-restart ?

Why don't you try docker? I run all the ROS stuff under docker so it doesn't kill my OpenCV library

I see, this was long time ago and nobody answered for some reason

please share the whole compiler output so I can help you out

to use rosmake it says ubuntu 12 and ros fuerte

You sure you have proper versions of ubuntu and ROS?

Are you sure you have good timestamps for tf? That's a common case.

yep. You sure that this timeout is not passed to the plugin? If not, then take a look at the service source code itself.