Alastair Toft @alastair@mastodon.gamedev.place

The main problem I have is that the noise computation is way too slow when generating the low LODs for the rest of the planet. I either need some much cheaper noise functions (maybe a 3D texture lookup) or to be able to do much fewer samples, or probably both.

Throwing a few different layers of Perlin noise on top of this already starts to make it feel like navigating a planet. Each biome has a random set of multipliers for each noise level.

Starting from a random distribution of points was causing some obvious problems. Firstly there's the chance of some very small regions, but also blending the regions together is more difficult when some points are bunched together - as the distance from the points to the boundaries is very inconsistent.

Starting out from a Fibonacci sphere of points seems like a good alternative - will just need to decide on a way to offset them from here.

Finally making a start on generating planets in my C++ project. I'm thinking of starting from Voronoi regions to define biomes 🤔

Ben Gibbard streaming Death Cab songs has been what I needed the past few days https://youtu.be/PX7A8tQNp8g

First time working from home today! Remote desktop is an interesting experience...

trying to make the view from your desk look like a snazzy home office and not just the one corner of your room that doesn't contain the bed or your laundry

It's on Linux too so that's neat

Playing Black Mesa this evening. Made it to the box smashing room

Started listening to this podcast and really enjoying it so far 🏳️‍🌈 http://www.mattbaume.com/sewers-summary/

Finally fixed some issues with the collision resolution in my C++ #gamedev project. If it taught me one thing, it's why literally everyone uses middleware for their physics code.