marvi wrote:Right now the whole process is a bit complicated for people not familiar with coding and it's not very general so I don't think it's a good idea.
But if anyone wonders, this modification didn't require 3D model editing (although it's possible to import any model to Blender). At the moment I'm writing a library allowing me to unpack/pack resources (written on C#, source code available on
GitHub). Using it, I wrote pretty straightforward script that tranforms original model coordinates, animations and a skeleton:
here.
By the way, I've updated res file, fixing wrong door attachment during opening/closing animations.
Thats just awesome!
But I did not get, why did you use a script to flatten cupboards?
Is it possible to decode any object model, update it in Blender manually and pack it back into res with your encoder?
If you'd be so kind to make short readme about how to use your encoder, it's can-do's and cannot-do's and maybe some points to pay attention. I'm coding in Java for almost 10 years and somehow understand C# so it may be as geeky as it needs to be =)
Wanna make smaller or transparent trees if I'll find some spare time.
2jorb: Sorry =) As soon as character will be able to find it's way around obstacles, I promise to use only original or beatified models =) Bumping into each and every tree during foraging is already an issue, but rotating camera to find out behind which tree my shroomz are doubles a hassle. As advice, A* algorithm with limitation to 1600-2500 square cells can solve the issue without killing performance
http://www.cokeandcode.com/main/tutorials/path-finding/http://arongranberg.com/astar/http://stackoverflow.com/questions/7355 ... -java-game