Update display if empty pointcloud2 is published #1073
Conversation
Do not show last point cloud any more, if published point cloud message does not contain any points
|
Thanks for the report and pr, I'll investigate it asap. If you have time to provide it, a simple bag file that reproduces the problem would make it easier for me to test this. As you can imagine, this is sensitive code (a break in this code would affect a lot of users), so I'll have to be quite cautious with it. If any other point cloud users can +1/-1 this I would appreciate it too. |
|
Sure, attached you will find a bag publishing some points on /point_cloud followed by an empty point cloud. |
|
Sorry it took me so long to get back to you. I've reproduced what you've described and that your patch fixes it. Thanks for giving a clear description and a reproducible bag, makes my job a lot easier 😄! I'll still do a code review before merging, but it's looking good so far. |
Do not show last point cloud any more, if published point cloud message does not contain any points
|
Backported in #1110 |
* Add fullscreen option. (#1017) * urdfdom compatibility (#1064) * Use urdf::*ShredPtr instead of boost::shared_ptr (#1044) urdfdom_headers uses C++ std::shared_ptr. As it exports it as custom *SharedPtr type, we can use the to sty compatible. This also adds a proper dependency for urdfdom-headers * adaptions to build against both urdfdom 0.3 and 0.4 ... relying on the compatibility layer of urdf package * Update display if empty pointcloud2 is published (#1073) Do not show last point cloud any more, if published point cloud message does not contain any points * Correctly scale the render panel on high resolution displays (#1078) * support multiple material for one link (#1079) * Fixed duplicate property name for Path colors (#1089) See issue #1087. * fix type error in newer versions of urdf (#1098) * Use unique material names for robot links. (#1102) * avoid C++11 feature for back port to indigo
Currently, a PointCloud2 display keeps showing the old point cloud if an empty point cloud message is published. I consider this a bug, since rviz still shows old data which is invalid in the meantime.