### Remarks and Errata

Some readers have requested clarification about the assumptions on the
capabilities of searchlights. In particular, about whether boundary
searchlights can rotate past their adjacent walls. As stated in the
literature, a searchlight is a rotatable beam sensor which ``cannot penetrate
the environment boundary''. It is intended here that the environment boundary
only restricts a searchlight's ability to sense, not its rotation. All
searchlights, even those located on the environment boundary, are free to
rotate 360 degrees. A beam may thus have length zero if its respective
searchlight is aiming into a wall, but it is free to aim there. This
assumption on searchlight rotation capabilities is the most natural because it
means that there is no discontinuity in a searchlight's rotation range as its
location approaches and then goes onto the environment boundary. There also
are important computational implications of the assumption, e.g., to guarantee
a rotation monotone schedule exists as in Conjecture 3.1 of the IJCGA paper
**A Complete Algorithm for Searchlight Scheduling** (Chapter 3, Conjecture
3.1 of the dissertation **Visibility Problems for Sensor Networks and
Unmanned Air Vehicles**).

In the GNC paper **Sampling-Based Roadmap Methods for a Visual Reconnaissance
UAV** and Chapter 5.3 of the PhD dissertation **Visibility Problems for
Sensor Networks and Unmanned Air Vehicles**, the convergence theorems
require a general position assumption. We provide a detailed description of
the general position assumption in Section IV.A of this
JGCD article.

Last modified: Sun Feb 5 23:07:27 MST 2012