Nikita Lisitsa

By popular demand, figured how to properly put the sandy river banks (aka point bars) on the correct (inward) side of the river! Now the meandering looks even more realistic with no actual simulation 😎

#indiedev #gamedev #indiegames

Fun world generation: a river almost made a full circle, enclosing a hex-shaped forest patch 🤯

#indiedev #gamedev #indiegames #screenshotsaturday

Just for fun, recorded a video travelling along one of the rivers from source to delta. Had to speed it up quite a lot, otherwise it was too long 😅

#indiedev #gamedev #indiegames

Just a random screenshot, standing on a hill and looking at the shore and distant forests

#indiedev #gamedev #indiegames

Oooph, fixed a bug in my wind field generation code, tweaked cloud density & spawning, and now they look and move soooo much better! 🤩

(this is sped up 256x, and I removed the day-night cycle for clarity; the video actually spans 14 in-game days)

#indiedev #gamedev #indiegames

River generation progress: fixed the "aqueducts" and weird mountain cutoffs, added occasional sand river banks. Still got a ton of issues to solve, though 😅

#indiedev #gamedev #indiegames

Having a flexible ECS engine underneath means I can debug clouds producing rain just by spawning rain "particles" (actually scaled & recolored boulders, lol)

#indiedev #gamedev #indiegames

Experimenting with particle-based clouds driven by the wind velocity field. Particles use scaled boulder meshes for test. The plan is to splat these into a texture which will be used to generate the final mesh as before.

#indiedev #gamedev #indiegames

This algorithm seems to be able to create insane meandering patterns with rivers stretching across the whole island, that's so cool

#indiedev #gamedev #indiegames #screenshotsaturday

Improved river height interpolation using cubic splines, fixed some weirdness on river banks & deltas. Still hunting a few remaining discontinuities...

#indiedev #gamedev #indiegames

@witchgirls@woof.tech Yep! I'm still not entirely sure about that, since now

let x = 10

• 1

and

let x = (10

would be entirely different pieces of code...

@oblomov@sociale.network I wanted to show a case where comparing to zero is the correct thing to do.

1. That's news to me. Do you have any references to support this claim? Sure, of your matrix is sparse/symmetric/tridiagonal/whatever, you can do much better, but for a general system my impression was that there are basically just GJ and QR. QR is more precise but slower.

Again, note that absolutely maximum precision wasn't the goal.

@oblomov@sociale.network 1. I don't see how this contradicts what I'm saying, except layer I show a case (slerp) where even FLT_EPSILON is not the correct thing, for a couple of reasons.

1. Absolutely agreed. Though note that the goal was not to have maximum precision, but to have reasonable results & nice invariants.

2bis. I've seen a lot of people who would immediately replace this code with epsilon, simply because of the "FP COMPARE BAD" mantra.