The date field's description read "Each moon's and dwarf planet's card says how far its orbit
strays from 1950 to 2100." Pluto is a dwarf planet on its card, but it moves on Standish's planet
elements, and the ETL measures only moons and the SBDB bodies against Horizons: its card ends
"Orbit: JPL approximate mean elements (Standish), fit for 3000 BC to AD 3000." and gives no stray
figure. In bodies.json, Ceres (7.2), Eris (0.1), Haumea (0.4), Makemake (0.3) and every moon carry
one; Pluto does not.
The description now reads "AD 1 to AD 3000, where the planets' and Pluto's elements hold. Each
moon's card, and Ceres's, Eris's, Haumea's and Makemake's, says how far its orbit strays from 1950
to 2100." The CLOCK_WINDOW comment and the scene's note comment make the same distinction.
Test: hud-dock 'opens the date field on the clock's date...' asserts the new sentence. Control:
putting the old sentence back fails that test only (1 failed, 836 passed).
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
The clock reaches AD 1 to AD 3000, but only the planets' cards named a span their elements hold
over; the 25 moons and the four SBDB dwarf planets gave a source and an epoch, though
BodyRecord.orbitSource is documented as "the span they hold over". And the worst offsets the ETL
stated came from twelve New Year's Days: Nereid's year is 360 days, so all twelve fell far from
its periapsis, where a mean ellipse is furthest out. The one date the ETL checked, 2025-01-01, saw
Nereid at 2.6 degrees; it reaches 11.19.
For each moon and each SBDB dwarf planet the ETL now fetches Horizons' ICRF vectors from 1950 to
2100, every other day (daily for Nereid, at an eccentricity of 0.75, and Hyperion, whose row's
eccentricity is a quarter of its real one: every other day gave it 22.14, daily 22.23), and
measures how far the mean elements stray, at the same TDB dates. The card appends it: "JPL SBDB
osculating elements, epoch 2026 Jun 9, within 7.2 degrees of Horizons from 1950 to 2100". Worst
offsets on the real catalogue: the Moon 2.62 (2010 March 27), Phoebe 2.58 (1969, where a comment
claimed "within 2.0"), Phobos 1.26, Mimas 7.43, Iapetus 10.34, Nereid 11.19 (2039 Nov 1),
Hyperion 22.23 (2055 Feb 26), Ceres 7.12 (1953); Io 0.07, Titan 0.06, Eris 0.06.
build.ts recomputes each from the same Horizons positions and fails if an orbit other than
Standish's names no span, if a card states less than it strays, or if a body passes its ceiling:
3 degrees, and Hyperion 23, Nereid 12, Iapetus 11, Mimas 8 and Ceres 8, each explained. The
2025-01-01 check stays for reading errors, its comment no longer passing one date's offsets off as
worst ones. The Sun's note says the moons' and those four's elements were checked from 1950 to
2100, and the date field's description that each card says how far its orbit strays over that span.
In the running app Ceres's, Phobos's and Nereid's cards end "within 7.2", "1.3" and "11.2 degrees
of Horizons from 1950 to 2100".
The CLOCK_WINDOW comment also had the calendars the wrong way at AD 1: proleptic Gregorian dates
are two days behind the Julian calendar there, level from AD 200 to 300, and ten days ahead by
1582. It now says so, and names Ceres's drift where it named Phobos's, which its orbit now carries.
Controls: the ETL measuring nothing fails ("Ceres's orbit ... names no span it holds over"),
rounding the stated figure down fails on Ceres (7.1 against 7.12), and Nereid held to the general
ceiling fails at 11.19; the note and the date field without the span fail their named tests.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Only setDate held the clock to AD 1 - AD 3000; julianDate did not, so a month a second carried it
past either end with nothing to stop it. Past AD 3000 it drew the planets on elements Standish
never fitted there, under a note naming "AD 3000"; before AD 1, toISOString writes the six-digit
years ECMA-262 uses outside 0000-9999, and the note, the date strip and the date field, which cut
it at fixed places, read "... to -000001-10-05 20 UTC.", "-000001-10" and an empty field.
julianDate now stops the clock at the end it ran into: re-anchored there, at real time turned back
into the window (forwards at AD 1, backwards at AD 3000), as if the reader had set that date. In
the running app, run backwards from 0001-01-10 at a month a second for 20 s, the note reads "to
0001-01-01 00:00 UTC." and the strip "0001-01-01", the clock 19.7 s into AD 1 at real time; run on
from 2999-12-01 for 8 s, "to 2999-12-31 23:59 UTC." at real time backwards.
Control: julianDate unheld fails "stops a running clock at either end of the window".
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
The clock from #33 could only run forwards from now. The Display panel's
clock now has a Date (UTC) field, a native datetime-local in a form, so Enter
submits it and the browser holds it to its min and max. It jumps the clock to
that date, and the clock carries on from there at the rate it was running at.
A Backwards toggle (aria-pressed) runs the same four rates the other way. The
radios still pick the rate's size and keep the direction when it changes.
The window is AD 1 to AD 3000. The end is where Standish's Table 2 stops
being fitted (3000 BC to AD 3000; every planet within 0.29 degrees of
Horizons at each date measured out to 3000). The start is the date input's
own floor. TimeStore.setDate refuses anything outside it, and NaN, and leaves
the clock where it was. The field is read as UTC. Its dates are proleptic
Gregorian, as a Date is, so before 1582 they run up to ten days ahead of the
Julian-calendar dates history gives. The window is written beside
CLOCK_WINDOW, with the moons' shorter reach (Phobos 11 degrees out by 2100).
The system note now names the date it is drawn for, to the minute:
"... to 2020-12-21 18:00 UTC.", or "to now, <date> UTC." at the present.
Measured in the app (port 4311, keyboard only: fill, Enter):
- Set to 2020-12-21 18:00 UTC, Jupiter and Saturn seen from Earth's drawn
position are 0.113 degrees apart. Horizons gives 0.102 geocentric
(geometric 0.1017, astrometric 0.1018). Distances: 5.9267 and 10.8296 AU
against Horizons' 5.9258 and 10.8270.
- 3001-06-01 is refused by the form (validity false) and the clock does not
move.
- Backwards at 1 d/s: -2.011 days in about 2 s.
- The field's accessible name is "Date (UTC)" and its description is the
window. Its colour-scheme is dark, so the picker icon shows on the HUD.
- Back to now puts the field back on the present too.
Unit suite 799 -> 805: two tests for the store, three for the dock, one for
the scene note. Eleven guarded mutants; each changed its file and made its
named test fail. Among them: the window check dropped, the wall clock not
re-anchored on a jump, the radio dropping the direction, the field read as
local time, and the note not naming the date.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
The off-screen rule for clicks was described in 3f0abf8 and in the pull request, but only its
comment was committed: pickAt still let the slop reach past the frame. The mutant that was said
to catch it matched nothing, and an unrelated flaky test failed instead. The frame test is now in
pickAt, before the slop, and its test fails without it (star at NDC 1.01, click at 0.995).
Venus, Uranus and Pluto turned forwards: Horizons states a retrograde spin twice, by a negative
rate and by an obliquity over 90 degrees, and both were applied. The period's sign is now used
only when no obliquity is known. Measured on the live markers, spin axis against orbit normal is
cos(obliquity) for each: Venus -0.999, Uranus -0.135, Pluto -0.494, Earth 0.917.
Moons listed as rates rather than "Synchronous" drifted about 5 degrees an orbit and Titan did not
turn: every moon is now locked at its Kepler period. Pluto's obliquity comes from IAU WGCCRE 2015,
Horizons gives none.
Also:
- the star's light is white at pi, not a warm 2.2 that left the photographs dim;
- procedural textures are 128x64, not 512x256 that froze the main thread ~60 ms a body;
- Io, Pluto, Titan and Deimos lose their "maps", which were disc photographs with black sky;
- an exoplanet with only a mass gets a radius from it (M^0.55, capped at Jupiter), not Earth's;
- the clock knows when it has left the present even once back at real time, so the date and
"Back to now" stay up.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Every body in the system view was an unlit sphere wearing a 32 by 16 pixel
procedural texture — the size chosen when a marker was a few pixels across and
what survived was its average colour. The thirteen real photographs in
`src/assets/textures/bodies/` were used only by the detail page. So Mars was a
pale grey ball with invented polar caps while its own NASA mosaic sat unread in
the repository, and nothing had a day side or a night side.
Each marker now takes its own photograph where one exists, at the size the
detail page uses, and the derived texture only where none does — the five moons
no probe mapped, and every exoplanet, none of which has ever been imaged. The
material is lit, and the light is a point at the star, so each world shows the
terminator where it really falls.
The light does not fall off with distance. Under the inverse square that real
light obeys, Neptune receives a thousandth of what Mercury does and reads as
black; the map is a set of worlds to look at rather than a light meter, so each
is lit as a photograph of it would be. That is the same concession the pixel
floor makes for size, and it is only about brightness: the *direction* is real.
Spheres are 32 by 24 rather than 16 by 12, since at true scale a body is drawn
anywhere from a pixel to the whole frame and the old silhouette was visibly
faceted at the near end.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The orbits and the rotations are both functions of a date, and the only date the
map ever asked for was this instant. So a view built on propagated ephemerides
showed a still picture: Earth turns 15 degrees an hour and takes a year to go
round, and a reader watching for a minute saw nothing move at all.
`TimeStore` is that date, at a rate the reader sets: real time, an hour a
second, a day a second, a month a second. It is read once a frame rather than
held in a signal — it changes continuously, and a signal changing sixty times a
second would ask the whole HUD to re-render for a number nothing is watching.
Changing the rate re-anchors rather than rewinding, so speeding up and slowing
down never jumps the sky, and "Back to now" returns to the world's own time.
Measured in the app, three seconds of watching in the Sun's system:
| rate | sky elapsed | Earth turned | Jupiter moved |
|---|---|---|---|
| real time | 0 | 0 | 0 |
| 1 h/s | 3.0 h | 45.12 deg | 0.0009 AU |
| 1 d/s | 3.0 d | (three full turns) | 0.0222 AU |
45.12 degrees in three hours is 15.04 an hour, which is Earth's own sidereal
rate, and Jupiter's 0.0222 AU in three days is its own orbital speed.
The rates are radio buttons, not toggles: they are one of four, and the native
control carries that to a screen reader and to the arrow keys with no script.
The date joins the strip only while the clock is running faster than the world,
since at real time it is today's, which the reader's machine already says.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
SenrokaiandClaude Opus 5.5