demo

Author: stevenctl

content/posts/tech/3d-autotiling/_index.md

@@ -2,32 +2,24 @@
 title: 3D Autotiling
 ---
 
-## Preface
-
-It took a few attempts to figure out how to organize this. At first it was
-written in the order of the things I tried out. I started with a very complex
-thing that led me to learn some more basic approaches. I will build from those
-foundational concepts up to the current state of my project.
-
-A while back I picked up a book on 3D modeling. It made me think I could make 3D
-games. Turns out there's a lot to it. And I got frustrated at my first attempt
-to kitbash some modular castle set into something decent. It didn't feel fun
-rotating and aligning everything properly.
-
-I got sidetracked for months learning a bit about procedural generation because
-I wanted a more interactive workflow. I implemented Marching Cubes in compute
-shaders, Heightmaps using compute and vertex shaders and my favorite thing: Wave
-Function Collapse with a simplified Marching Cubes. 3D autotiling was the sweet
-spot I was looking for.
-
-This project has exposed me to a bunch of areas in the graphics and asset
-creation pipeline. I'm pretty happy with where it is right now and I'm hoping I
-can polish it off as a solid tool for making small and big games. The ultimate
-goal is a level of polish that can be exposed to end users to create their own
-levels. I don't think I would have become interested in shooters if it weren't
-for Halo 3's forge mode. I probably logged 1000 hours in Minecraft when I was a
-kid (and learned Java because of it!). Games that let you make stuff are the
-best.
+<video src="demo.mp4" autoplay muted loop></video>
+
+I picked up a book on 3D modeling and then tried to start making a game.
+Building a castle felt a bit tedious to me, even with a modular kit. In 2D and
+3D, I hate rotating and aligning tiles to fit. 
+
+My impatience led me down a months-long rabbit hole learning about procedural
+generation. I've come up with what a pretty decent Wave Function Collapse
+implementation, including Blender tools to help with asset creation. I
+implemented the core algorithm 4 times. First in Blender/Python intending to
+export to Unity/C#. Unity was annoying so I tried Bevy/Rust. Turns out an
+editor is a valuable debugging tool and I finally have a solid working version
+in Godot. The ultimate goal is a grade of polish with which end users can
+create in-game or in-engine levels.
+
+I don't think I would have become interested in shooters if it weren't for Halo
+3's forge mode. I probably logged 1000 hours in Minecraft when I was a kid (and
+learned Java because of it!). Games that let you make stuff are the best.
 
 ---
 
@@ -41,17 +33,25 @@ the gaps if I explained things well enough.
 {{< cards >}}
     {{< card
         link="01-marching-squares"
-        title="Part 1: 2D Autotling with Marching Squares"
+        title="Part 1: 2D Marching Squares"
+    >}}
+    {{< card
+        link="02-basic-wfc"
+        title="Part 2: Basic Overview of WFC"
+    >}}
+    {{< card
+        link="03-driven-wfc"
+        title="Part 3: User Driven WFC"
+    >}}
+    {{< card
+        link="04-generating-tiles"
+        title="Part 4: Generating a Tileset"
+    >}}
+    {{< card
+        link="05-sockets"
+        title="Part 5: Simpler Adjacency"
     >}}
 {{</ cards >}}
 
-## First Drafts
 
-The following are brain dumps that will be rewritten.
 
-{{< cards >}}
-    {{< card
-        link="01-marching-squares"
-        title="Part 1: 2D Autotling with Marching Squares"
-    >}}
-{{</ cards >}}