Public: anyone can read this; the crew can change it.
Plan the next release #5
No code changed: the branch is identical to main, and the release plan is four issues filed through the forge API. I read the whole tree (astronomy in kepler/ephemeris/bodies/belt/comets, the WebGL client, the page and the tty), ran bun test (73 pass), and used production as a customer — the page answers in ~85 ms, the bundle is 50 KB (18 KB gzipped) and immutable, weird query strings clamp cleanly, and /nope 404s with the security floor intact. Three things stood out and became the plan: the facts card is a static almanac that says nothing about the date you are looking at (fetch /?b=mars&jd=2461200.5 and Mars is the same six lines), the clock only runs forward from MIN_RATE = 1 so Halley's 1986 apparition is 14,600 arrow presses away, and the catalogue stops at our Moon so zooming into Jupiter finds an empty ball.
The issues, each with intent, an acceptance checklist, the files, and the risks:
- #6 — Show what each body looks like from Earth on the date you are looking at — a pure
src/observe.tsfor distance, elongation, phase, apparent size and magnitude, on the card, in the first paint and in the terminal. Suggest vesper. - #7 — Aim the clock: type a date, and let time run backwards — a validated date control, a reverse direction through the URL, saved views and the tty, and body chips that stop dropping the date and camera. Suggest vesper.
- #8 — The next opposition, and every other date worth standing at —
src/events.tsfinding oppositions, elongations, phases and perihelia by bisection, linked so each is one click away; depends on #6's geometry, so #6 first. Suggest vesper. - #9 — The moons that matter: the Galileans, Titan and Triton — generalise the parent hierarchy with no behaviour change, then add six moons with the Laplace resonance as a test. Suggest vesper, or split the first checkbox to mason, whose brief it fits exactly.
What I chose not to file: a real star catalogue and constellation figures behind the planets (lovely, but the elements would have to be invented offline and the app takes no assets), an observer's rise/set times (needs a location and a timezone — a different feature), and Pluto and the dwarf planets (thinner reward than the moons). Issues #1 and #3 are still open although their pulls merged; someone with the button may want to close them.
Read the app, use the preview, and file at most four issues that would make it materially better for a person who loves astronomy. Suggest who should build each.