After just under a fortnights progress, I have finally got around to making a 720p video of my WIP Hex Engine, covering a few of the features that are in there right now…
Cheers normen, good idea, the filters are hard coded in but wouldn’t take much to put some controls in Nifty.
Thanks Skye, the music in the video starts during one of the peaks in my mix… which was a bit dumb… I kinda just slapped it over the video… which reminds me: for anyone who cares enough, you can grab a copy of the full 30 min mix here.
Was watching it with some VFS classmates. Our jaws dropped when you removed the underlying piece to make the castle. I just assumed this system would have similar restrictions to a heightmap, so this was a very pleasant surprise.
@erlend_sh said:
Where are you going with it from here?
The engine will form the base for the next game I'll be working on, I plan to build that game as a mod for this engine... not sure if this will work yet but it has really helped to better structure my code i.e. keep game specific logic out of lower level systems.
@nehon said:
How many time did you take to make this??
I've been working on this for just under 2 weeks, full time, with a few days off for "research" ;)
I was designing a Hex Engine to handle the terrain of my game. Units wouldn’t be hex-based, only the terrain itself, maybe some decoration and buildings. And on the top of that, I could easily build a navigation mesh for the AI. But that’s all on design, charts, images and pseudo code yet
Sorry to ask, but do you plan releasing the engine?
I don’t plan on releasing the source any time soon if that’s what your asking.
The game details are still being worked out, what I can tell you is it will be hex tile based, there will be building, and there will be some pathfinding
Currious: Is that path finding feature in you game → a “Hex A* path finding” Algorithm ?? Look like the most advanced AI stuff I saw in the last few years, since Hero might and magic!
@atomix more or less, similar to A* but with the addtion of a weighted algoritm bias variable that can be set between 0 and 1… 0 being Best First Search, 1 being Dijkstra, and A* falling halfway between the two at 0.5f (it’s kind of a search speed vs optimal result type dial)… running a basic euclidean heuristic for now, then multithreaded and wrapped neatly into an appState;
Now for as wanky as all that sounds, it’s actually pretty simple, this stuff is will work with any shape tiles or nav meshs, as long as you have ‘nodes’ (or ‘tiles’) and you can calculate their neighbours and distance between them, the rest is a bunch of fancy sorting… more or less.