Chassis, dimensions and costs.

Family comparison

Body radius differs by family even when bodySize = 1. Actual radius scales with the cube root of actual biomass and inverse cube root of dish volume. Depth thickness is computed separately; cilia, flagella and glow are decorative. The solver uses ellipsoid clearance in XYZ and swept contact.

Core expression and geometry

These rules are shown in their code units. A registry gene’s relative value enters one or more equations; it is never a temporary energy or body-mass store. View focus and zoom are observer settings.

What controllers decide

Consumer sensory updates run at 200 ms intervals and retain last-seen positions for up to 1800 ms. Threat, contact, satiation, cooldown, eligible local prey, carrion and recent sighting compete in that order in grazer.js:stepGrazer. Hunger goals differ by family. The sensor records local observations; the observer may see more of the dish. Chemical concentration does not reveal a source path, and blocked approaches can be abandoned.

Algae sample local light and mineral probes every 200 ms, mark the limiting input, then turn toward improving samples or explore. Prototype chassis use separate locomotion and resource paths; some remain development seeds. See controller metadata for source symbols and null values where no unified priority rule exists.

Resource flows

Environment defaults and bounds →

Environment control ranges

These are UI input bounds and initial control values from index.cfm. A loaded dish may contain different current settings. All space, mass and energy units are artificial. Camera focus and zoom do not alter biology. The control table is a reference, not an import contract. New dish feature defaults: chemicalCues: on, contactExchange: on, fusionPropagules: on, acquiredDna: on, toxicEffects: on, algaeCommunication: on.

Temperature is °C; volume is a relative multiplier; other percentage sliders are 0–100 unless identified by their own unit. Cycle period and fade inputs are seconds in the UI and milliseconds in state. New dishes use automatic nutrient additions by default; current pulses are off until enabled. See environment defaults for the state representation.

Reproduction thresholds

Fission first checks family support, population and support-species caps, age, 1.4× mature body mass, fuel and recent algal surplus. It then searches up to 16 physical daughter placements before spending the 3-energy division cost; parent body and remaining fuel split between daughters. Mutation is sampled independently per locus in each daughter.

Contact exchange requires two distinct mature algae with crossAffinity ≥0.5, at least 50 fuel and 1.2 seconds of actual XYZ contact, with no contact gap longer than 240 ms. Each parent spends 2 fuel and keeps a partner genome commitment for a later daughter.

Gamete parents require fusionCapacity ≥0.7, crossAffinity ≥0.5, age ≥30 s, mass ≥1.15× mature mass + 0.20 and fuel ≥30. They spend 0.20 mass and 10 fuel for a 45-second gamete. Opposite mating types (matingType <0.5 versus ≥0.5), different parents, five-second age and XYZ encounter within 0.02 physical units are needed for fusion. The zygote recovers eight seconds before ordinary development.

Crawler spores require sporeInvestment ≥0.7, age ≥30 s, mass ≥1.15× mature mass + 0.25 and fuel ≥32. A release spends 0.25 mass and 12 fuel for a 90-second asexual spore. Germination needs five seconds of viability, local organic field ≥0.002, physical room, a cell slot and support-species slot. The shared particle cap is 128. Expiry deposits remaining particle mass to the nutrient medium. These mechanisms require the Fusion/propagules feature toggle.

Chemical channels

These are field packets with distinct channel and group identities. Concentration samples are local, divided by dish volume and gated by receptor sensitivity. A receptor detects a gradient, not emitter identity or a stored path. Paid release transfers body mass plus 20 energy per mass; toxin release transfers a previously synthesized store. Passive food and rot odors transfer accounted material without active secretion energy. Packets advect with current, random-walk, widen and decay into nutrient material. The 256-packet cap merges compatible nearby packets where possible; other release is blocked before charging the source.

Detection threshold is 0.00000015/max(1,volume) for food and rot, and 0.000001/max(1,volume) for other channels. Toxin dose, resistance and clearance are physiological processes after field detection. Individual response may be neutral even when a sample registers.

What this reference cannot predict

No static genotype determines a net energy or survival score without local food, movement history, light, encounters and competition. Contact shapes approximate bodies; an elongated cell does not have a deformable mesh. At the default mutation rate 0.12 per locus, a 46-locus alga expects 5.52 attempted perturbations per daughter; unchanged or clamped results reduce the changed-locus count. Looping scent trails, moving obstacles and depth routes can defeat simple pursuit. The catalog’s habitat label describes a usual setting, not a viability guarantee.