[1.16.4] TopSolidRangeConfig topOffset mystery

[frakier]

I tried tracing the code but that went off the rails pretty quick.

what does topOffset do?
some samples from the minecraft code...
public static final ConfiguredFeature<?, ?> ORE_GRAVEL_NETHER = register("ore_gravel_nether", Feature.ORE.withConfiguration(new OreFeatureConfig(OreFeatureConfig.FillerBlockType.NETHERRACK, Features.States.GRAVEL, 33)).withPlacement(Placement.RANGE.configure(new TopSolidRangeConfig(5, 0, 37))).square().func_242731_b(2));
   public static final ConfiguredFeature<?, ?> ORE_BLACKSTONE = register("ore_blackstone", Feature.ORE.withConfiguration(new OreFeatureConfig(OreFeatureConfig.FillerBlockType.NETHERRACK, Features.States.BLACKSTONE, 33)).withPlacement(Placement.RANGE.configure(new TopSolidRangeConfig(5, 10, 37))).square().func_242731_b(2));

the official minecraft wiki has this for the generation...

Veins of blackstone attempt to replace netherrack 2 times per chunk in veins of size 1 to 33, from levels 5 to 28, in all Nether biomes except basalt deltas.
Veins of gravel attempt to replace netherrack 2 times per chunk in veins of size 1–33, from levels 5 to 41, in all Nether biomes except basalt deltas.

So the times per chunk makes sense... EX[func_242731_b(2)]
The 33 makes sense... EX[33] easy enough.
TopSolidRangeConfig falls apart, a bit but then the site is based on 1.16.2 it seems.

So does this sound right?
(5, 0, 37) would be 5 to 37
(5, 10, 37) would be what 5 to 47... 15 to 47? ??

Edited December 21, 2020 by frakier

[TheGreyGhost]

Hi

public class RangePlacement extends SimplePlacement<TopSolidRangeConfig> {
   public RangePlacement(Codec<TopSolidRangeConfig> codec) {
      super(codec);
   }

   public Stream<BlockPos> getPositions(Random random, TopSolidRangeConfig config, BlockPos pos) {
      int i = pos.getX();
      int j = pos.getZ();
      int k = random.nextInt(config.maximum - config.topOffset) + config.bottomOffset;
      return Stream.of(new BlockPos(i, k, j));
   }
}

 

Beats me why they found it necessary to have both config.maximum and config.topOffset in there.  Probably it means something, eg maximum is the highest y value of this feature but don't use the topmost topOffset rows.

 

TopSolidRangeConfig(int bottomOffset, int topOffset, int maximum)

means (5, 0, 37) gives you nextInt(37  - 0) + 5 = 5 to 41 inclusive

and (5,10, 37) gives you nextint(37 - 10) + 5 = 5 to 31 inclusive

 

-TGG

[frakier]

On 12/21/2020 at 4:46 AM, TheGreyGhost said:

Beats me why they found it necessary to have both config.maximum and config.topOffset in there.  Probably it means something, eg maximum is the highest y value of this feature but don't use the topmost topOffset rows.

TopSolidRangeConfig(int bottomOffset, int topOffset, int maximum)

means (5, 0, 37) gives you nextInt(37  - 0) + 5 = 5 to 41 inclusive

and (5,10, 37) gives you nextint(37 - 10) + 5 = 5 to 31 inclusive

Could be. That makes sense compared with how they do do the range(32) which is basically 0-31.
I get a chance I will do some testing with some xray and see how different settings work out.
Figured I would see if someone knew the answer before I started too far down the rabbit hole.

Edited December 23, 2020 by frakier

[secknv]

Did that testing, the conclusion was exactly what TGG said (well since he got it from the code that would be obvious lol):

// bottomOffset -> minimum height for the ore
// maximum -> minHeight + maximum = top level (the vertical expansion of the ore, it grows x levels from bottomOffset)
// topOffset -> subtracted from the maximum to give actual top level
// ore effectively exists from bottomOffset to (bottomOffset + maximum - topOffset)