Ehh... I'd a kind of... rather make you you think that it's not worth the effort...
(to go back to nice simple shortest path).Ok, ok. That would be unprofessional...
Likely the best I can do is to mark fire tiles as not walkable and ignore them - what do you think?
I think it sounds as a good solution. Because, if I understand what it entails, it means that tiles under fire won't be highlighted in any way, thus it also clearly communicates to players that those tiles aren't considered for the pathing at all.
However, it needs to stay in tune with the path. So then if clicked on tile on fire -> mark tiles on fire as walkable? Though, not bullet-proof.
Or, alternatively, what about a shortcut to be hold, or even better, a shortcut (+button) to switch them both between two states: (1) walking around hazards (default) and (2) the simple shortest path -> letting us choose regardless a tile clicked? Uff, asking for more work. But perhaps less stress on the mouse-tracking input system, as tiles on fire are either simply ignored in case of (1), or not in case of (2)).
Different algorithms are used for pathfinding and displaying reachable tiles. The pathfinding algorithm is a pretty efficient implementation of A* - even so it has to do quite a lot of work to get a path. If you click on a tile, it gets 1 path. If you think about what's happening with the display - it could require the finding of 80+ paths. So instead, I used a less work intensive method of getting the data.
That's interesting. Vaguely thought display is more demanding, but hadn't actually considered how many combinations you need.
Probably not useful in any way: From the openxcom code I recall, they first try to calculate path using fast Bresenham algorithm and if can't find any, only then A* is used. Obviously, it only helps to optimize if Bresenham doesn't fail often. X-COM maps also have more open space compare to yours.
Noticed that the game often shows path, even when we aren't allowed to go due to insufficient AP. I understand it may be useful for future turns. Just musing how it could differ when pathing would be calculated and displayed only using tiles within AP radius.