Before each snapshot I like to do some playtesting in survival, just to check that things look OK. Several people have suggested I record such a session, so I figure why not? Here's a playtest session from the morning before releasing experimental snapshot 6. I also explain a bit about how the new terrain generation works, including the new stuff on the F3 overlay.

So if you like the idea of watching a dev wander around in minecraft for an hour and rambling about terrain generation and design stuff, and trying not to die in the process, then this might be for you :)

You can download the experimental snapshot here:
https://www.minecraft.net/en-us/artic...


0:00 Intro
2:56 Continentalness
10:22 Erosion
11:07 Temperature & Humidity
13:38 Erosion and Mountain Placement (Tibet)
17:58 Peaks & Valleys, Weirdness
20:24 Peaks & Valleys and Saddle Valley generation
22:38 5D noise + unpredictability
26:15 Challenges of microbiomes, no minimum biome size
32:14 Weirdness, Intervals and the selection of biomes
34:36 Floating blocks
35:29 Erosion values and cliff formation
39:58 nearWater: Aquifers and coastlines
42:07 Water caves, dry caves, barriers
43:22 MC quirks and the "door trick"
44:47 Carvers and dead ends
47:32 2D vs 3D noise maps
48:44 New vs Old MC generation (tradition)
49:22 Continentalness, Erosion, coastlines, and Swamp placement
51:11 Factor: Shattered Savannahs and Extreme Hills
56:09 Old 1.17 3D variation (mostly forests)
56:53 The challenge of fixing rivers vs land traversal
57:55 nearWater: Aquifers and rivers
59:22 The necessity to code Aquifers
1:00:35 Aquifers interacting with rivers, realism vs gameplay
1:02:05 Trouble with coding Badlands, esp. without minimum biome size tech.
1:03:54 Choice of "let them be crazy" with respect to crazy big cave entrances
1:04:51 Deserts and Badlands, Intervals + microbiome "Oases"
1:10:02 1.17 vs 1.18 Badlands