Thanks a lot for the help vtrix. If I understand you correctly, I can make many different invisible objects, no matter they will use many different low size textures, not combined in an atlas, as long as i keep them hidden.

It makes sense.

anyone?

I guess that all depends then... For 20x20 pixels sprites 300 times instanced, its better to use a 512x512 sprite i guess. But for bigger resolution you will need quite big map for all trees, and its a hell of a loss of pixels.

I might be wrong, but that's how I understand it.

Its always better to use many instances of an object, than to make a big map with the tree duplicated many times.

Hi guys

Recently i noticed something, I'll need your help with.

I noticed that for blank objects, I mean fully transparent, used only for detecting collisions, on export i got black PNGs exported. For now I'm making them 2x2 pixels, and then scaling them to the right size. I wonder is there a way not to use maps at all, or all transparent objects to use only 1 map, 2x2 pixel?

And one more thing. Is there a way to make 2 or more different sprites to pack their maps in 1 atlas? I know if sprites are different animations or frames of a single sprite object, they'll pack, but can i do it with different sprites?

Thanks

Very nice example. Thanks a lot!

I don't quite understand you about joining the arrows but looking at the example i guess something like that would do the trick:

Find path to (Sprite.Pathfinding.NodeXAt(NodeCount), Sprite.Pathfinding.NodeYAt(NodeCount))

Hopefully, it would give the position of the last node of the path.

Cant wait to go home and test it

Yeah, that was my first thought too, storing coordinates. I was hoping, there was a better solution tho. As i see it, a Recalculate Path action for the pathfinder behavior would be a nice addition.

I believe there should be a way to determine where is the last node of a path, giving you the destination coordinates.

Hi guys

I'm making a RTS like game and as any such game I plan to have units walking here and there. The problem is when I place another building, and regenerate obstacle maps, those units that are on their way needs to recalculate the path based on the new grid, right? But i don't have a clue where they are going... Is there a way to learn their destination so i can recalculate the path? I hope i will not have to store the coordinates in variable or something...

Great, thanks.

I was setting most of my events to be formula based, thinking that is cheaper, but i prefer using "on collision" and "is overlapping" conditions.

Ashley

Is there a way to use only the cheap collision tests based on the bounding box, without using the actual polygon check after. Or maybe the 4 point poly checks is as fast as the simple collision check?

I was always wondering how much better is to use simple distance check like (Obj1.X - Obj2.X) < 100 ; (Obj1.Y - Obj2.Y) < 80, than using collision checks?

There is a Unit update if anyone is interested :)

Still needs alot of polish but the base is done.