towards a faster A*

so i’ve finally begun the A* implementation for Caveman 3.0.  i first played around with A* almost a decade ago. this time around, with my typical outside the box approach to things, i decided to try things a little differently. to begin with, i’m working with a 300×300 grid type search area. once the code is complete this will most likely be changed to a variable sized search area. instead of open and closed lists, i keep an A* map: a 300×300 array with f,g,h, and state: open | closed | neither. for the moment, a separate 300×300 collision map has the obstacle data in it. but the two maps might be combined. the only slow part is getting lowest f, which requires searching the entire grid. so the idea…


Link to Full Article: towards a faster A*