Addressing the limitations about csa-challenge

[espadrine]

This project is meant to remain simple. However, I believe it can be done by keeping track of overall travel duration, similarly to more classic pathfinding algorithms.

I am not sure whether the tests currently require not to have that bug fixed, but it is possible. This project is only for fun, to have a go at comparing language and implementation performance!

[linusnorton]

Yeah I was just wondering what the most peformant way of doing it would be. I thought about continually replanning at later time periods until the original arrival time is no longer achievable or running an inverted version of the algorithm to find the latest departure that still arrives by the original arrival time.