I spent some time recently just watching players hit the opening of the game and reading feedback about early experiences, and it’s been illuminating.

Nothing is fundamentally broken, but players are getting in their own way, and that’s on me.

So rather than layering more on top, I’ve been going back and reworking the first 10 minutes with everything I’ve learned over the past months.

The biggest shift is in the tutorial.

One of the best critiques I've heard is that the first thing you learn in the library is not the first thing you do out in the wild. The library feels safe and forgiving. Stepping into the world adds stakes, and that shift can be overwhelming while you’re still learning how to interact with everything.

Most of the early friction is clarity, not difficulty. Players need to feel comfortable interacting with the world before they’re asked to solve problems inside it. There are a few surgical moves I can make in the tutorial to set the table better before the feast begins.

I have plans.


Over the last couple weeks I took a serious run at reworking structure borders.

The goal was to shrink the borders in an attempt to clean up visual noise and isolate their silhouette better. When structures are tightly packed, it can be hard to read each one cleanly. In practice, this change cascades into a huge amount of art rework edge cases, so I’m shelving that direction.

The exercise wasn’t wasted though. It revealed a handful of specific culprits (like grindstone salt shards and simple gripper fingers reading too similarly), and there are meaningful gains to be made by targeting those directly instead of attempting a full art rework.

This is one of those cases where the ideal solution doesn’t survive contact with reality, but the exploration still produces a better, more focused plan.


On the lighter side, I added birth animations to Saltways and Streamways.

Small change, but it does a lot of work. These are foundational logistics pieces, and having them smoothly slide into the world instead of popping in improves both readability and feel. The animation also reinforces travel direction, which helps immediately.

It’s a bit of polish that shows up in the first 10 minutes and carries through the entire game.


Another major axis of first impressions is environmental detail. The terrain shader is strong, but it’s missing some ambient depth via pebbles, grass, seashells and the like.

I took a swing at GPU instancing for this and ran into issues with MoteMancer’s specific rendering setup (split 2D/3D, heavy stencil use for portals). It’s not playing nicely there.

Not done with it yet though. I have another approach in mind that borrows from how terrain borders work, which might get a similar result without the same constraints. I’ll come back to it after the tutorial work is further along.


The biggest holistic takeaway is that players will tolerate hitting edges, but that’s not a reason to leave them in.

MoteMancer opens up once you get past the learning curve. My job is to make sure the path there is smooth and readable enough that players actually reach it.

More soon, back to the lab 🌿

~CyanAvatar