Fifteen bodies here have a real photograph. Every exoplanet does not, and
never will on current instruments — none has ever been imaged — and nor do
several of the solar system's own moons. Those all shared one crude
stand-in: a few noisy bands tinted by category, cached per colour, so
every exoplanet in the app was literally the same picture.
They now get a surface reasoned from what has actually been measured.
The chain is standard at every link. A host star's luminosity comes from
its catalogued apparent magnitude and its parallax distance — that pair is
exactly an absolute magnitude — plus a bolometric correction for its
spectral class. The correction is not optional: an M dwarf radiates most
of its light in the infrared, so its visual magnitude understates it more
than tenfold, and M dwarfs are what most nearby planet hosts are.
Luminosity and the semi-major axis then give an equilibrium temperature,
mass and radius give a bulk density, and size, temperature and density
together give a class of world.
Checked against the solar system the temperatures land on Earth 255 K,
Jupiter 112 K, Neptune 46 K, all within a kelvin or two of published
values, and 51 Pegasi b comes out at 1227 K against a published 1200.
Each class carries a palette reasoned from its chemistry — methane absorbs
red light, which is why the ice giants are blue — and a structure: zonal
bands for a body with a fluid envelope, because a rapidly rotating
atmosphere organises into them, and fractal terrain for one with a solid
surface. Polar caps grow and shrink with the derived temperature, which is
the clearest visible consequence of the whole chain.
The generator samples three-dimensional noise along the sphere rather than
a flat field, so there is no seam to stitch at the antimeridian and no
pinching at the poles, and it writes into a byte array rather than a
canvas — a pure function, testable, with no 2D context to be unavailable.
Two things the derivation cannot do, both stated on screen next to the
measurements it rests on. Equilibrium temperature ignores greenhouse
warming and internal heat, so Venus comes out at 300 K against a real
surface of 737 K and Io, kept molten by tides, classifies as ice. And
these are illustrations: reasoned, but not observations.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WaySiNst4HhDXBHnMy8p5G
The map stopped at the catalogued 50 pc around the Sun — 0.33% of the
Galaxy's width — and looked like a point cloud with a search box.
Adds the galactic scale above it and the heads-up display the reference
map is built from.
The Galaxy is not a third coordinate space. It is the same parsec space
four orders of magnitude further out, so the model and the star field
crossfade against camera distance instead of switching, and the Sun stays
where it really is: 8.18 kpc out, on the Orion Spur, between the
Sagittarius and Perseus arms. The depth range scales with that distance —
one fixed near/far pair cannot both fly into a star and hold the Galaxy.
The structure in shared/astro/galaxy.ts is measured: the directions of the
centre and the north galactic pole, which fix the disc's 63 degree tilt
against the celestial equator; the Sun's galactocentric distance; and a
radius, azimuth and pitch angle per arm. The particles scattered around it
are not, and cannot be — dust hides the disc, so no catalogue holds the
Galaxy's stars. The view says so, and the model fades out before the
camera reaches the 50 pc where the real stars are.
The rest is the look: polar grids lying in the galactic plane with drop
lines from the Sun's neighbours, a scale ladder, a readout panel, range,
reticle and frame brackets. Two things had to give way for it. The
deep-sky shell is the sky as seen from here, so it dissolves rather than
letting the camera fly through a wall of nebulae, and so does the skybox,
which is a photograph taken from inside the thing now being viewed from
outside. Labels are picked by screen separation rather than distance
alone: the Sun's fifteen nearest neighbours are all inside four parsecs
and printed as one unreadable clump.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WaySiNst4HhDXBHnMy8p5G
The design doc scopes deep-sky objects as a galaxy-view backdrop and lists
fetchDeepSky.ts, deepsky.json and deepsky.model.ts, but none of it existed —
it was the only part of the plan with no implementation behind it.
ETL: fetchDeepSky.ts pulls the OpenNGC catalog, classifies each object as a
galaxy/nebula/cluster, and keeps the ~460 worth drawing (everything Messier,
everything with a common name, and anything brighter than magnitude 9) out of
~12,000 mostly-anonymous rows. build.ts runs it and validates the output.
Distances are the hard part: OpenNGC has no distance column, and both fallbacks
fail for the best-known objects. M31, M33 and M42 are Local Group members whose
redshift is negative or absent, and a galaxy's catalog parallax comes from a
cross-matched foreground star — 6 mas for M31 would put a 780 kpc galaxy at
167 pc. So records store a unit direction on the celestial sphere rather than a
position (the line of sight is always known precisely, and the objects are drawn
on a fixed backdrop shell where true distance is unusable anyway), and distance
is optional metadata carrying its own provenance. Parallax is trusted only for
galactic objects, redshift only above z=0.003 where expansion outweighs peculiar
velocity. 330 of 463 get a distance; the rest honestly report none.
Rendering: DeepSkyRenderer paints the objects as soft additive billboards on a
2500 pc shell — clear of the 50 pc star field, beyond the camera's 2000 pc orbit
limit, and inside its 5000 pc far plane. Size comes from real angular extent, so
Andromeda is six times wider than the full Moon, clamped at both ends. Sprites
rather than points because the WebGPU backend caps point primitives at one pixel;
materials are shared per kind and brightness band, so 460 objects cost nine of
them. The brightest dozen get permanent labels, which needed the label overlay to
accept string ids alongside numeric star ids. The backdrop is decorative, so a
failure to load its dataset is logged and the star field comes up regardless.
Also documents the app in the README, which until now covered only the plugin
marketplace.
Tests: 112 passing, up from 54. Build, both typechecks and the Playwright suite
are green.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WaySiNst4HhDXBHnMy8p5G
Add star-map Angular app, ETL pipeline, and caveman plugin
Angular 3D star map (galaxy/system/body views, Three.js rendering,
navigation store) plus the NASA ETL tooling that builds the star,
exoplanet and solar-system datasets, Playwright e2e suite, and the
cs:caveman Claude Code plugin (command, agent, skill).
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
@
Claude
Senrokai