[Devlog] March 2026 - Week 4

Welcome to the 68 week devlog for Cosmic Construct. As I often mention I have alot of side projects I upkeep, and have commitments outside of developing the game. This turned out to be one of those weeks so I didn't get much done. (I've also been playing Phantasy Star Online private servers. ;--; It's just part of who I am.)

To follow the game's progress you can join me at:

• My bluesky account which gives more frequent updates.

• The Cosmic Construct community discord server.

• Also Wishlist Cosmic Construct on Steam and follow it's steam page.

---

Pawn AI

Despite not working on the AI much this week, I did manage to get the pawns largely to a state where they are again running around performing their tasks which is a good sign. I also did a much needed backup of the development files. The last backup I did was in October of 2025 *cringe*. (Although, technically the code is updated much more frequently on github.)

Side Projects

Other than maintaining "the Keygemin IP" as a side project, I occasionally feel compelled to write chapters of the stories set in that universe. I honestly feel terrible for leaving my books on a year long hiatus while I go off doing other things. I do wish I was able to set off to finishing one task at a time. Even within the scope of writing fiction novels, I write multiple books concurrently. Is this healthy behavior? No. I do finish my projects eventually. :P

I've been testing a brand new web fiction publishing platform, by the name Fictionry. The features of which I really like. Needs some polish, but I think it could, even in its beginning stages, have real potential to compete with larger platforms. I saw that immediately coming in as a user and wanted to help the developer there.

So if you are a budding author that wants a good chance to be seen on a platform that had potential, look there as an opportunity to stand out. Better if you do the writing yourself, if you know what I mean.

I've done a bunch of rapid prototyping for some 3D game dev projects. The screenshot in front of you is one of the failures of that prototyping, in an attempt to procedurally generate a dungeon from component parts. The columns you see here are supposed to be on the corners of the platforms holding them up, and snap to origin issue. I decided its better to use prebuilt rooms and socket them together. I'm not sure if I'll even use this in the future, but I have alot of ideas in my brain that I need to see realized to feel satisfied about. Once they are realized, I can check them off a mental list of what to use for my "theoretical ideal game". My fellow gamedevs should know how that feels. --- I'll say pawn AI again, regardless if that actually gets done. Development is slow sorry about that, I'm just one guy. - Pawn AI