Hi everyone,
We have some news to talk about before we proceed, so lets get that out of the way first
First off, We've unfortunately had to move our launch date to the 19th of December, this was due to some misunderstandings regarding steams internal publishing rules, fortunately this will give us a little more time to do some final testing before launch.
Secondly, We've managed to get Steam achievements integrated into the game much sooner than expected, so the game will be launching with steam achievements!
We have 23 achievements at launch, with a few on the way for each content update we release! This includes two secret achievements that you'll have to go Really far out of your way to get.
Here's a few icons for achievements you can get throughout the game.
Anyway, lets talk about todays main topic the Pea pathfinding AI.
Let's use an example, can you solve this incredibly complex maze?
Well, a computer would struggle with this.
Pathfinding is a very tricky concept for computers to understand, since they can't see ahead the same ways as us. People have put lots and lots of effort into making the perfect pathfinding algorithms that work bug free, and more importantly, quickly.
The most common Pathfinding technique is called A* (A Star), it works by breaking the world down into nodes, then we try to draw a straight line towards the tile until it hits a wall. we then go one tile at a time around that wall until we find the correct way, this will result in the shortest path between the start and end points.
Of course there's a but more to it than that, and each game will have some other steps is does to match it's specific needs, but that's the very basic explanation of how it works.
if you've ever played a game (especially a strategy game) with large maps and wondered why the Game slows down or acts funny when you try to move units over long lines, it's probably because it has to check thousands and thousands of nodes to find the best path for you.
as you can see this has lots of times it goes to the wrong path, which would waste valuable processing time, especially if we had lots of Peas doing this at once.
we take a slightly different approach, a bit different called a Wavefront Expansion Algorithm.
Instead of going directly to the end point, we start at the end point, then go in all directions, until we've been everywhere in the map,
by keeping track of which way we went for each node, we end up with a big bunch of lines that always show the way back to the end point, the computer tower.
you might have noticed that both of these methods give us the same path, which is a good sign that we're doing this right.
because these lines cover the entire map, we can then have a Pea simply follow the line to the guard tower regardless of where they are.
since all they're doing is following a line, which uses much less processing power, we can then have a lot more Peas on the screen at once, it doesn't even matter where they start, since the lines will always show them the way to their shared goal, the tower.
That's all for now, See you next time!