Wolfgang Hönig

Could you include the command line you used to generate the output from the input?

(This is the mapf/c branch). From a CBS perspective, both results are actually identical (and optimal), because the cost is the sum of the timesteps of all robots. If agent0...

What is the cost for self-loops then? For example, your graph has edges `['6', '6']`, so if we minimize path length, 6 -> 10 -> 5 has the same cost...

Also, should the path length be just a tie breaker? For example, is `1 -> 2 -> 3 -> 4` (3 timesteps) better than `1 -> 5 -> 4` (2...

Yes lets discuss tomorrow! I have implementations for multiple versions. The tie-breaking variant computes what you wanted in this particular example. The other variant (with a little hack to avoid...

The version with the new cost function (min path length and count self-loops like the shortest edge in the graph) is now available in the issue40 branch (see https://github.com/whoenig/libMultiRobotPlanning/pull/41 for...

This is an example for a single CF. If you plan to fly multiple, you have to change the example or use the crazyswarm (https://crazyswarm.readthedocs.io/en/latest).

This is definitely a supported combination. Two things come to mind: 1) You seem to run a fairly old firmware version. I am not sure when the getDeviceTypeName query was...

Which branch are you using and do you exchange any other packets apart from the setpoints? How do you measure "very poor performance"?

I mostly work with https://github.com/whoenig/crazyflie_cpp/tree/dev-crazyflie-link-cpp, which uses an entirely different way of communicating. Would be great if you could try and report back if that solves your problem. I assume...