Thanks guys !
We’ve been 18 months at this, with a team of two. It’s going slow because there is so much stuff to do, but it’s getting there.
Well space is mostly empty so we base navigation on collision avoidance, not on pathfinding. Ships go were they have to (next dock, next target…) and avoid obstacles in the way. For every object, we compute potential trajectory intersections, and how far they are. If a trajectory intersection is near, we thrust in the direction opposite to the other object’s trajectory. It’s a basic algorithm, works well in 2D or 3D.
A complicated part was to handle the ship itself, since we have 20 to 30 individual thrusters on every ship - all contributing, all destructible.
Thanks ! No, we keep to Slate and C++. We’re both C++ programmers at heart so Slate was fine, UMG didn’t even exist when we started, and it’s still not interesting for us today.