Expanding the perimeter by one tile in cardinal directions helps get rid of the single-tile "protrusions" seen on the far left side of the previous screenshot (those tend to be awkward/more likely to require a guess, which is less-interesting gameplay).

Still need to think about what all this means for coordinates when playing with TTS (don't think I'll do anything special, but there's a big amount of "wasted" coordinate space in the top left and bottom left).

Because these are generated, I get a bunch of variation for free. I don't think that variation feels like "content," but it helps make digs feel less samey, which is a big deal for the sort of game this is.

I've still got a bunch of other stuff to wrap up before I can start focusing in earnest on content and variation density, but it's exciting to on the cusp of.

I mocked up some museum floor textures for the last video, and variation there from such a minimal change is more than I expected.

I spent some of today's stream poking around with hotloading meshes and attaching them to existing skeletons (for character customisation).

I already do this in places with skeleton fossil chunks, and I'd done some previous experiments, but it seems I'd overlooked that vertex group information seems to get broken/stomped on if meshes are exported separately - and not consistently (if they're *imported* as separate files, that's fine). A little frustrating, but I think I have something workable.

@Cheeseness Wait there's a time slider?

That's what GPlates https://www.gplates.org/ has! I used to work on it!

@neko Very cool! I have GPlates installed here. What sort of stuff did you do?

Fossil Sweeper's time slider was one of the first things I worked on when I started this larger version. It represented a significant unknown that I didn't have a workflow for, so it made sense to front-load (here's the first video of this version of the project that I recorded back in January 2022 :D )