Ores: Remove overlapping regions. Make some regions deeper

[paramat]

Remove overlapping coal/iron ore regions for consistency with all
other ores.
In lower coal/iron regions increase number of clusters to compensate for
removed overlaps.
Decrease dimension of these clusters to better match the 'ores to cluster
volume' ratio of the small clusters.

Move density increase for coal and iron to y = -127.
Make tin deeper with region depths identical to copper.
////////////////

Previously for both coal and iron:
The top region of small clusters continued down to world base, overlapping the deeper region of large clusters. This is from early MTG which had few ores, now we have many more making ore generation more intensive. This PR removes this overlap but preserves the ore density in the lower region by making the large clusters more common.
Removing the overlap will make ore generation a little less intensive.
The small clusters of the upper region is unchanged so this doesn't affect gameplay for beginners.

Previously the lower regions for coal and iron started at a much too easy y = 0 and y = -64, move the lower region down to start at y = -128.

Make tin deeper, since it is equivalent to copper in the progression of ores make the depths of its regions identical to copper: upper starts at y = -128, lower starts at y = -256.

This PR is therefore the second part of the 'make ores deeper' task started in #1813 In that PR i didn't alter coal and iron regions, so this PR does that.

The ore progression of value and depth is now (same line = same value/depth):

Coal, Iron
Tin, Copper
Gold
Diamond, Mese crystal
Mese block

[Ezhh]

This causes over 200 coal ores in one place at any depth under 0 (tested in game to be certain). This is far too high.

[paramat]

Good point. Will change.

[paramat]

I suggest still doing what this PR does but for the lower distribution: Instead of the same number of clusters and them being much larger, have clusters of the same size but more of them?

[Ezhh]

As an example for reference, crazy amounts of iron at only -200:

The wall directly behind me also had a comparable amount.

I'm not opposed to more clusters at lower depths, just opposed to large clusters like this (probably at any depths).

[paramat]

PR is a mess but my new suggestion for the lower distributions keeps cluster size almost unchanged (it has to change a little from to keep ore abundance unchanged) and increases cluster number.
For both ores previous clust_num_ores was 27 and clust_scarcity was 24 * 24 * 24.

	minetest.register_ore({
		ore_type       = "scatter",
		ore            = "default:stone_with_coal",
		wherein        = "default:stone",
		clust_scarcity = 12 * 12 * 12,
		clust_num_ores = 30,
		clust_size     = 6,
		y_min          = -31000,
		y_max          = 0,
	})

	minetest.register_ore({
		ore_type       = "scatter",
		ore            = "default:stone_with_iron",
		wherein        = "default:stone",
		clust_scarcity = 12 * 12 * 12,
		clust_num_ores = 29,
		clust_size     = 6,
		y_min          = -31000,
		y_max          = -64,
	})

The distance between cluster centres has halved from 24 to 12, so there are 8 times more clusters.
Thought i'd check with others first before adding this to the PR.

[Fixer-007]

As an example for reference, crazy amounts of iron at only -200:

It is occasional cluster, sometimes it is crazy low amount of iron. I do agree that in MTG iron/coal is VERY-VERY abundant, in caves it is bonanza.

[Ezhh]

What I saw when testing wasn't occasional. That was the first cave I found, and had two large clusters. There was a third in a tunnel joining it, and a fourth just a little further on.

I find myself questioning why we need to raise the amount of common ores. I understand people expect mining deeper to give more reward, but something that doesn't have actual need/use isn't really rewarding.

Can we please add worth to mining deep by having actual content there?

[paramat]

It is occasional cluster

In my first version of this PR (which is what Ezhh was referring to) there were 7^3 clusters separated by 24 nodes (centre to centre not edge to edge), so not rare.
In MTG master the clusters are 6^3 separated by 24 nodes.
My suggestion above is 6^3 clusters separated by 12 nodes, but we could make the clusters denser (4^3), as is more realistic.

If we preserve the amount of coal in the lower distribution we cannot avoid large clusters that are fairly close. Coal tends to be found in large seams so i think large clusters is more suitable than lots of small ones.
WIll update the PR to my latest suggestion.

[Ezhh]

I'm really against large clusters and against overpopulating the underground with common ores. We do not need "8 times more clusters".

[paramat]

I find myself questioning why we need to raise the amount of common ores.

There maybe a misunderstanding, to be clear, this PR will not increase the amount of coal or iron ore at any level.
Because the contribution of the small clusters is now added to the large clusters there has to be more large clusters, and 8 times more is what is needed to maintain the total ore amount. However there are now no small clusters.

For coal (iron is very similar):
In MTG master in each 24^3 node volume there is 27 small clusters and 1 large cluster.
In this PR, once updated, in each 24^3 node volume there will be no small clusters and 8 large clusters.

[paramat]

See #2047 (comment) my suggestion does not increase the size of large clusters, only the number of them.

[paramat]

Also we can reduce the dimension of the large clusters if desired without reducing ore count per cluster.

[paramat]

Updated but working comments are left in for the moment for my use.
Also reduced cluster dimensions to 5^3:
"Also decrease dimension of these clusters to better match the 'ores to
cluster volume' ratio of the small clusters."
If ores per cluster is to not rise this is how it has to be.

[Ezhh]

I can try some testing over the weekend, but this honestly feels like a case of fixing what isn't broken. It also sounds like it's reducing the tiny bit of variation we do already have underground. I know you want to reduce the number of decoration registrations active within any given map height, but the more I think about it, the less I like it.

Would appreciate some input from others on this one.

[paramat]

Yes i can understand some may like having 2 cluster sizes. It's mostly for simplicity and consistency with other ores. We will add underground variation other ways.

[paramat]

@sfan5 @SmallJoker @rubenwardy any opinions?

[SmallJoker]

Overlapping ore definitions are certainly not optimal as they result in too high ore densities. I haven't tested the PR but the concept looks good.

[paramat]

The overlap doesn't cause a too high ore density because it's always been this way. It's just a case of consistency, simplicity and being less intensive. I agree that having 2 cluster sizes in a volume is 'nice' but despite that i'd prefer consistency with other ores. Having 2 cluster sizes is a leftover from mgv6 days when we had very few ores, i think we can modernise and move on from that. Researching coal and iron deposits IRL, large clusters seem realistic.

[Ezhh]

By "large clusters" how many ores per cluster are you intending?

[paramat]

The PR is updated with my new suggestion. The large clusters are smaller in size, have roughly the same number of ores, and have closer spacing. Overall 'ores per node' density is unchanged.

[paramat]

Here's a suggestion: Coal has its density increase start at y = 0, we could move this down to y = -64 to match iron, then no more large coal clusters just under the surface, the small clusters would go down to y = -63. Coal is too easy to get currently.

[paramat]

I could also make tin like copper: start at y = -128, density increase at y = -256. I see tin and copper as similar in value, and are combined for bronze.
I'll update soon, WIP.

[Ezhh]

I mainly just want to make sure ores per cluster isn't too high for any of them. :)

[Fixer-007]

Hmm, could you also increase rarity of ore clusters in general? Ore seems super abundant to me. Like/dislike

[paramat]

Fixer instead it's better to make the y of ore density increase lower so that roughly the same amount of ore is present in a world but you have to go deeper for it. I already did this recently but could tune again later.
I've spent enough time altering these clusters, i think they're optimum now, will make some y changes.

[paramat]

Note this branch is called 'moreores' but that is actually not the case.

[paramat]

Updated to make some ore regions deeper as suggested above, first post rewritten to explain all changes. I'm fairly happy with this now.

[paramat]

Messed up branch attempting to rebase so reopened in #2056