Magical gene / changes a rule, paints nothing

Extreme white dominant

The only gene in the mod that changes a coat without painting a single pixel. While it is present, a white texel is final: every other gene still runs, in the ordinary order, and every one of them is thrown away wherever the horse is already white. On a horse with no white it does nothing at all. On a horse carrying both a white marking and something magical, it swaps which of the two is on top.

What it does to a horse

A horse's white markings normally go down first. Tobiano, splash, sabino, a blaze, a stocking — all of them are natural genes, and natural genes resolve in phase 1, before any magical marking exists. So a horse carrying both a tobiano and something loud comes out looking like the loud thing: the magic is painted over the top of the white and the white is gone.

Extreme white dominant reverses that for one horse. The white patches stay exactly as the melanin genes drew them — same shape, same edges — and the magical marking fills in around them, stopping dead at the boundary. A galaxy horse with a blaze keeps the blaze. A cleave horse with four stockings keeps the stockings.

It is white that is protected, not the natural genes

The rule is about the colour, not about which gene put it there. White a magical gene paints is locked from the moment it lands too, so a marking that goes down white is protected from every marking above it in the order. That is what makes it “extreme” rather than “high priority”: there is no slot in the paint order a gene could take that would get over it.

The eyes are the one exception

Eye colour is left alone. A depigmented blue eye is pigment biology rather than a marking, and the eye texels of a heavily white horse are whitened along with the rest of the head — so locking them would delete every eye-colour gene on exactly the horses with the most interesting eyes. Nothing else is exempt.

Crossing two of them

One copy is the whole of it and two look identical: “white is final” has no second half to reach. That makes it a genuinely dominant magical gene, which is unusual here — most of the invented loci are recessive, so that finding one is a breeding project rather than a catch.

Finding one

About one wild horse in a hundred and sixty carries a copy, and because a single copy expresses, that is also roughly how often you will see one — but only on a horse that has white and a magical marking to argue with, which is a great deal rarer than either. Most horses carrying it are indistinguishable from horses that do not.

Its gene carrot

Gene key
horsegenetics.extreme_white_dominant
Class
ExtremeWhiteDominantGene
Phase
magical — but it has no painter in either phase
Priority
a code-order slot only; see below
Alleles
n EWD
Outcomes
wild, white-dominant (a marker(), not a wild type)
Capability
WhiteLockContribution — exactly one implementor
Coat effect
none of its own; it masks every other gene's
Built, never seen in game

No horse carrying this has been rendered in a running client. What is worth looking at first is the edge: a magical marking stopping at a tobiano boundary should read as a hard cut, and if the two disagree about where the boundary is it will show as a one-texel fringe. Checklist on To be verified.

Why it is a capability, not a priority

The obvious way to write this gene is to give it priority 999 and repaint the white last. That is not the same thing, and the difference is the whole gene. A priority is a slot: it says "apply me after these and before those", and a marking sitting above the slot paints over the white exactly as before. There is no number high enough, because "highest" is a position and the rule is not positional.

So it is WhiteLockContribution instead, which the composer reads out of band — the same shape LutContribution uses for the gradient swap. The composer asks, once, whether any gene on this horse locks the white; if one does, the answer is a mask, and the mask is applied at every step of phase 3 and again over the overlay pass. Its own PRIORITY is therefore only where its segment sits in a genotype code.

How the mask is built

  1. Seeded from phase 1. After the natural genes have finished restricting pigment, every texel left with essentially no pigment at all (red and black both under TRANSPARENT_EPS) is locked. Those are exactly the natural white markings: phase 2 leaves them fully transparent, and what shows there is the bare white template.
  2. The eye rectangles are cut back out — see the gameplay tab.
  3. Grown after every magical gene paints. Any texel the accumulator now holds as near-white (every channel at or above WHITE_MIN) joins the lock. Near-white rather than exactly #FFFFFF, because a gene paints white with a TOWARD at 88–98% strength and lands a few counts short on every channel; an equality test would lock the natural white and silently miss all of the painted white.
  4. Read by ColorField.apply(delta, locked), which skips locked texels entirely — absolute paint and additive tint alike — and by CoatOverlay.paint, so an overlay gene cannot get round it either.

A null mask means no gene asked, which is the ordinary case; every caller downstream takes a nullable mask rather than an all-false array so a horse without this gene costs nothing.

Its place on the paint order page

It does not appear on the paint order at all, because that page lists genes that paint and this one does not. That is the honest answer and it is also the point: the stack is not where this gene lives.