0 capacity of a single vehicle available should lead to no route

[rsynek]

Having a single vehicle a decreasing its capacity to 0 does not remove a route.

Expected: similar to having no vehicle at all, no route should be shown

[yurloc]

Interesting point. Technically, vehicle with 0 capacity can still take the route and visit places. It just can't deliver any packages.

The model should allow this. This is what a ghost vehicle would do in an over-constraint scenario (as a replacement for nullable variables, that don't work with chaining).

From frontend point of view we're just visualizing backend data, if OptaPlanner assigns a chain of visits to a vehicle with zero capacity, we have to represent it somehow. Probably indicate very clearly that the vehicle is unable to satisfy the total demand of that route.

Not a bug but interesting discussion point. Closely related to #151.