Integrated Path Planning Object (T0) Version 0.2

Here is the latest build (0.2.2) of my new pathfinding object. This is intended to be a very efficient path planning solution (for huge maps) that is also easy to use.

Functionality that is in the "Under Development" menus are unstable, so don't report bugs about them.

Feel free to report bugs here, PM them or email them to me. If you will be working a lot with this object then you may want to use this Please login to see this link. to see what's already being worked on and report more complicated issues.

All comments are welcome, not just bug reports. If you have suggestions that can make using it easier or increase its utility, be sure to let me know.

Glad you liked.

Porting it to Flash will depend largely on the state of affairs at the time it is complete. I would really like to have it available for Flash of course, but I'm not sure how easy/hard that would be. One thing that I may do would be to build a slimmed down version for Flash that is compatible but not as powerful as the original one, but no promises on this front yet, as my ActionScript experience is minimal.

Alright, let me know what you think.

Looks great, will use it in the next version.

Thanks

Here is version 0.2.3.

Now using Pharanygitis' icon (although it seems to have lost some pixels when it was resizes to 32x32).

Fixed Ross's bug as well as the shaking bug where it fails to detect that a unit is at destination.

Here is an example that includes both direct object manipulations as well as classical grid pathfinding. For the grid pathfinding I have put my object side by side with the two other pathfinding objects by z33z and Ross, to serve as reliable benchmarks for the new object. Unfortunately only Ross's object reports time consumed in search (as far as I know) so I couldn't display meaningful statistics yet.

This example uses Looki's Surface object, so you will need to install that as well if you don't already have it.

Quote from SortaCore

With this map:

I can't see this map, URL is broken. Can you upload it here (or somewhere else) or to my issue tracker?

It sounds strange that the APF object reports failure where the PF object finds one.

The IPP object is probably just taking much longer, if you leave it running for a while it should eventually terminate. I believe that no possible path is available since this is the worst case scenario, but I'd like to see your map.

Right now there isn't a failure mechanism yet, but that's planned.

Oh, there's also a small bug in the IPP which I already fixed but didn't bother to upload. Basically it shifts the map a bit (i.e. if you put the start/destination adjacent to blocks it may believe they are ON the block).

But since you ran into this, here is the fixed version.

I think I should implement saving/loading for this example to help with reproducing bugs.

OK, I can see this one. Let me know if you can reproduce it with the new version.

No, the 16 is a virtual size. Each block you draw visually is represented by 16*16 actual pathfinding nodes. The pathfinding is all done on the 1-pixel tiles, which is why you see it as a line instead of bunch of large squares.

I'm using a grid size of 1 to stress test the objects to stimulate huge maps.