Lucas WintherandClaude Opus 5.5
50c6e91b00
The new reader app replaces the old UI: Checklist with a rail (focus pills, Next to expire, today's dailies, pinned routines), Timeline in a new and a classic look, Routines, Archive and a Settings area, in a Structured (default) and a Relaxed interface. It follows the approved Round 6 B and Round 7 mockups, corrected screen by screen with the owner. Reader data moves to a v2 store (gacha-tracker:v2:*). Existing v1 data is migrated on first load, a recovery copy is written first, and the v1 keys are never removed, so a rollback finds them unchanged. Redeem codes for Genshin arrive as their own feed. Also: live's urgency ramp, editable per theme; completion and section-complete fireworks; a returning-user tour; the design mockups, specs, plans, reviews and handoff record. Co-Authored-By: Claude Opus 5.5 <[email protected]>
Snapshots
Raw pages, exactly as fetched. scripts/refresh-sources.ts writes them; nothing else fetches.
<source-id>.html the body verbatim — tracked
<source-id>.meta.json hash, size, charset, ETag, Last-Modified, contentChangedAt, lastConfirmedAt — tracked
<source-id>.state.json transient run bookkeeping: last attempt and failure streak — gitignored
"Verbatim" means the bytes as served, not text we re-encoded. charset in the metadata records
what those bytes are in — the Content-Type header, else a <meta charset> in the page, else
UTF-8 — and bytes is the served length. A page in Shift_JIS or Latin-1 decoded as UTF-8 would be
stored as a field of U+FFFD with the original bytes gone; re-parsing could never recover it, and the
mojibake would reach slugify, moving every event ID for that source (AGENTS.md § Event IDs are
localStorage keys).
Every file is written to a sibling .tmp-* and renamed into place, body before metadata, so an
interrupted run leaves a stray temp file rather than a truncated snapshot or metadata describing
bytes that were never stored. Those temp files are gitignored: refresh.yml commits the whole
directory, so otherwise a run killed mid-write would pin a half-page in git forever.
Three reasons this is committed rather than cached:
- Re-parsing never re-fetches. Iterating on a parser reads these files, not the wikis (AGENTS.md § Scraping conduct).
- A refresh is reviewable. The commit diff is the page diff, so "an event vanished" is a question you can answer from git rather than from a wiki that has since changed again.
- The build stays offline.
bun run build:feedparses whichever of these exists and falls back tofixtures/otherwise, so a clean checkout and the container build work with no network.
The .state.json files track transient check attempts and failure streaks. When a source is
confirmed (200 or 304), lastConfirmedAt is recorded in <source-id>.meta.json so downstream builds
(including GitHub Pages, which deploys from git without access to the Gitea Actions cache) know the
exact time the source was confirmed current.
The metadata is also rewritten on an unchanged page when the server rotates an ETag or
Last-Modified while serving the same bytes. Keeping the old validator would mean sending a stale
If-None-Match forever and being served the whole page every cycle, so that diff is worth the
commit — contentChangedAt and the body stay put, so it is still visibly not a content change.
Fixtures are not the same thing. A fixture is pinned to a date and kept forever as the regression test for a page shape; a snapshot is the current page and is overwritten each time it changes.