It's not possible just out-of-the-box, unfortunately.

The issue is that typical physics engines assume no border and no wrapping. The physics engine you need should be fixed to allow for wrapping, which would also impact the way collisions are handled.

Chipmunk and Bullet, the engines we have at ...

What didn't work for you when you used gdb?

You do it in the loop functions, not in the robot controller. You reimplement the loop function method "IsExperimentFinished()", as shown in the embedding examples. No robot can single-handedly stop the simulation.

In your code, the use of print is not correct. It seems to print a double using the format for an int.

Also the return statement in the same function contains a redundant assignment.

Did you notice the position of the light sensor on the kilobot,, and how that affects the calculation of the readings? Also the position of your light is at height zero, which is unusual.

The answer to all of your questions are in the original Kilobot paper and in the source code. I encourage you read both, they should make everything clear.

what is the unit of measurement of the value specified in the <light> tag in the intensity attribute?

It's an intensity multiplier. See this ...

You found a bug, but fortunately there's a workaround. When you initialize the LED medium, try this:


<led id="leds" grid_size="2, 2, 2" />


Essentially, what is happening is that the grid size to store the LED entities is calculated as 0 because your environment is smaller than 1m x 1m. It's a ...

Great! Yes, that makes sense.

Yes, you should use the point, and then maybe get a different error if something else is off.

It's just a typo. Exercise for you to fix it