Replicating was easy. Add the Djinn to Alorra, then click somewhere far away. They move a few tiles, then click done. They continue all the way to the end (no intermediate steps this time, maybe because of only 1 Djinn?).
Edit: Hm, if you add Alorra to the Djinn instead, they act normally.
The move certainly does get interrupted when the Djinn used up its movement and starts to become carried by Allora but they can't move more than what they are supposed to afterwards - the remaining moves are correct.
Unless your stack ended up further away than 6 tiles from where you added Allora (if she was the last one) or 4 from the last Djinn (if you added that), there is no bug.
Pretty much this is how it goes : Djinn moves to Allora, now has 2 moves. You send the stack away. As the top unit, the djinn, can move 2, the stack moves 2 then stops. Allora loses 2 moves, but still has 1. Djinn would lose 2 moves - but cannot : The stack still contains a wind walker with more than 0 moves, so it is set back to 0.5. You can then move 1 more tile and that's it. It Allora was fully ready, you could have moved 3 more. If you picked up a fresh Djinn on move 5, you could have moved another 4 from that. Etc.
Huh, I don't understand this at all. I add the Djinn to Alorra and click like 30 tiles away (or however many I feel like, as they have infinite movement at that point), all the way across the ocean. They move two tiles, some fuckery happens, then move the other 28 (but you can only do it once, as they sync up after that).
So they definitely end up way further than 4/6 tiles total.
Tried again...it seems to only happen if the second move is an "automove" caused by the done button. If you do it any other way it does not have the bug. I wonder why...
Well it doesn't seem to be directly wind walking related. The code for that is correct, if the wind walker does not have movement left after moving, the ability won't trigger to keep the exhausted units at 0.5 moves.
Which means, the only possible explanation for this is, that movement does not subtract the movement cost somehow.
While investigating for the possible cause of an AI stack not attacking,I discovered that the real unit rating calculation and the tool that calculates these are not in sync. More specifically, the tool outputs double the actual value for all units.
This isn't bad in most cases - the relative strength of units is the same either way.
However for one thing it matters. The limits set on how strong army the AI needs to be able to attack are set to twice the intended amount. This affects attacking with weak stacks and affects Jihad. Basically, while the AI was supposed to be able to attack with any stack being at least as strong as 2 halberdiers, it instead needed 4. This greatly reduced its capacity to strike at small settlements with a single swordsmen in them.
I still feel like high men don't really need crusaders and knights. They don't seem to fill a unique enough roll in the highman lineup, highmen are already fine later on with paladins, I do agree, earlier on they can be quite underwhelming. I had this idea while watching hadriex's latest caster of magic video session.
How about a high man gladiator or fisherman or something? Give them the spider's ability of web? have them require I dunno fighter's guild? and/or shrine? make them a bit like a replacement for crusaders? slow and fairly weak? probably 4 figures? would that be OP?
Then again, I'd like to see these new unit slots you freed up in the code used more on races I feel need the loving even more, like say gnolls....
