Rainbow dust

Crossing two of them

It paints nothing

All three outcomes are wild types, so affectsCoat() is false, the locus stays out of a horse’s texture key, and the genotype gallery collapses it to a single entry — the same trick particle, milk and verdant use.

Its gene carrot

Gene key

horsegenetics.rainbow_dust

Priority

151 — immediately after the particle locus

Alleles

Rbw n

Outcomes

wild, rainbow-dust-carrier, rainbow-dust

Shows when

Rbw/Rbw only

Default allele

n

Wild frequency

0.25% of founders are Rbw/Rbw; no founder is a carrier

Founder draws

1 × nextFloat()

Coat effect

none, ever

Particle

dust_color_transition, anchored at hooves

Emission

while moving on the ground, 50% of ticks, 2 particles a firing

Epigenetic

one value — how fast the colour turns, 40–300 ticks a lap

Recessive, with no wild carriers

One copy shows nothing at all, exactly as at the particle locus and for the same reason: a trail is a thing you breed for.

The founder table follows the same rule too. A wild horse is either plain or the whole rainbow — never a carrier — so what you catch is what you watched it do, and the carriers appear one generation later, in the foals of a rainbow horse bred to an ordinary one. The argument is written out in full under the particle locus’s founder table.

How the colour turns: the cycle parameter

An emitter carries a fixed color and color2, which is right for every other user of the verb and useless here — a rainbow is a colour that is a function of time. So the verb grew one parameter:

record Emitter(String kind, String shape, String anchor, Trigger trigger, int color, int color2,
               int count, double data, String particle, double chance, int cycleTicks,
               Condition when, int minDose)
        implements GeneAbility {}

cycleTicks is 0 for every other emitter in the mod, and means “the two colours above are the colours”. Anything larger is how many ticks one full rotation of the hue circle takes, and the translator then recomputes both colours from the clock on every firing and ignores what the gene put in them — color2 a twelfth of a turn ahead of color, so the dust_color_transition fade reads as “red going orange” rather than as two unrelated colours in one puff.

It is a parameter rather than a second verb because everything else about a rotating trail — the particle, the anchor, the count, the chance — is the emitter it already was, and a whole verb whose only difference is where two ints come from would be a copy of this one. Any data-driven gene can now say "cycle": 120 and get a rainbow.

The hue is counted off the horse’s own tick count rather than the world’s, so two rainbow horses standing side by side are not in lockstep — the same reason the healer’s aura beats on its own clock.

The one thing that varies

How fast the rainbow turns, in ticks per lap, drawn on the allele copy and inherited with it. A horse near the fast end (40 ticks — two seconds) strobes; one near the slow end (300 ticks — fifteen seconds) looks like a solid colour that has changed by the time you next look at it.

It is one number rather than a colour, and the colour is deliberately not the gene’s to vary: the rainbow is the gene. But two rainbow horses side by side should still not be the same horse, and this is the smallest thing that makes that true. It is a SCALAR, so drift moves it by a hair each generation and a line really can be bred faster or slower — which is the only thing about this gene there is to breed.

Why it is not a forty-first particle allele

It is the obvious question. The particle locus already has forty alleles, every one of them “this horse trails something”, and adding a forty-first is one line.