Okay, I just have to get attention: I know there's been some threads about this, and also that there is (at least should be) snippets in the codebase regarding A* and other pathfinding algorithms, but I just had to figure this on my own. For some odd reason. And now that I've done it, I need some publicity, otherwise it would've been coded for nothing.

...and I thought I'd also ask at the same that if people could go over this and give feedback (how to improve, bugs, etc...).

I know about some already, but it's in enough working condition to say that it works. 8-D

I believe that most of the overhead in that is basically the "live observation"...

Here are some media files for it (1bpp bitmaps representing maps)
http://corrie.scp.fi/users/jani.karkkainen/m100x100.bmp
http://corrie.scp.fi/users/jani.karkkainen/m8x8.bmp

Agh. HUGE Bug(tm). A little bit of thinking and some more reading on the subject, I believe it is now bugless (hah! ). Here's the new code.

Umm. Damn.

Third time is the ticket. This time I won't put the whole code, it seems too wasteful.. so, on line 125:

if (estimate < curnode.cost + curnode.heuristic)

change that to

if (estimate < a_list(closed, inclosed).cost + a_list(closed, inclosed).heuristic)

and NOW it should work correctly. If it doens't, I'll curl up and die.

It works ... if you change the mapwidth/mapheight variables to match your bitmap size

Nice one!

hehe, i had to wait abit but it works good!

Oh yes, I noticed that too... I'll be fiddling with it a bit for a while, and probably tidy the code and stuff, before I incorporate it into my game... I think it would be way faster if I discarded the drawing of the map and progress and so on. I'm not gonna test that yet though, because it's fun to watch that thing go!

Thanks for the input, gives me more motivation to continue to code!