Revision history [back]

Caution speculation on my part: I believe the difference to scan_time/ranges.size() is supposed to account for the dead time, when e.g. a 240 deg laser is making a full revolution. The time increment in this case is probably 240/360*scan_time/ranges.size().

As to rviz: This is likely due to the fact that transformations now have to be computed for each individual ray instead of once per scan. However, if algorithms can run "more correct" I would not stay away from this feature due to the visualization, but rather fix rviz.