rqt_robot_monitor: tree does not go stale when diagnostics stop

[ahendrix]

When diagnostics are no longer published, the diagnostics tree that is displayed should reflect this.

The old RX dashboard did this by graying out the dashboard panes if no diagnostics were received within 10 seconds.

You can reproduce this behavior by playing back any bag file with diagnostics data, and observing the dashboard behavior for 10-15 seconds after the bag file finishes.

[ahendrix]

This is also a problem on startup; the old robot monitor would be grayed out and display "no message received" on startup.

Old robot_monitor when stale:

rqt_robot_monitor when stale:

Old robot_monitor on startup:

rqt_robot_monitor on startup:

[ablasdel]

The knowledge that nothing has been received seems like rather important information. This could alert the user to a connection issues that could be serious and that they shouldn't trust the robot_monitor info anymore.

[ablasdel]

This should be a rather simple addition of a timer/color change. I will look into getting this done.

[ablasdel]

This will be the new look when the diagnostics timeout. This color indicator will be in addition to the bottom clock indicating time since the last message was received.

[ablasdel]

These will fix the issue on groovy/hydro. I would like someone to check my Qt logic before we merge these.