- renderPaste now looks up the paste's attachment and passes it as
'Attachment' so the paste.html attachment chip/image preview renders
- remove duplicated .iconbtn.dark-toggle blocks (3x from stacked
branches #127/#128) and duplicate mobile-media variants
Verified: go vet + go test pass; live render on :18098 shows
attachment-chip + img src=/f/id/name for a png upload (served 200
image/png) and chip-with-filename, no preview, for an html upload.
- Collapse the 10 theme cards into 5 pairs: light swatches top row, dark bottom row
- Dark mode toggle in settings and topbar (sun/moon inline SVG)
- Toggling dark mode switches to the other variant of the selected pair
- Persist pair id in localStorage palette-theme, dark flag in palette-dark
- Head script resolves base+dark to variant; URL ?theme= accepts both ids and wins
- Settings gear highlights like nav tabs on /settings
- PALETTE_DEFAULT_DARK env var sets server default dark state (default on)
- Server-side resolution helpers + tests
- Collapse the 10 theme cards into 5 pairs: light swatches top row, dark bottom row
- Dark mode toggle in settings and topbar (sun/moon inline SVG)
- Toggling dark mode switches to the other variant of the selected pair
- Persist pair id in localStorage palette-theme, dark flag in palette-dark
- Head script resolves base+dark to variant; URL ?theme= accepts both ids and wins
- Settings gear highlights like nav tabs on /settings
- PALETTE_DEFAULT_DARK env var sets server default dark state (default on)
- Server-side resolution helpers + tests
t.colors was removed by #113; rendering threw on first iteration and
the theme grid rendered empty. Use the colors returned by
presetColors(t.id) instead. Verified in headless chromium: 5
theme-card elements render on /settings.
Titles are trimmed and truncated to 200 characters; language must be
blank or match ^[a-zA-Z0-9+#-]{1,40}$ (max 40 chars) or the create is
rejected with a clear 400. Regression tests cover truncation, malformed
and oversized language values, valid identifiers, and blank metadata.
GET ?password= and X-Paste-Password header checks in handleGetPaste were
unlimited (pentest2: 8 wrong = 8x401, no 429). Gate every password
verification on the same per-IP+paste unlock limiter (5/min, 429 with
Retry-After) already used by the POST unlock form path.
Regression tests: repeated wrong passwords via GET ?password= and via
header must eventually 429; correct password within burst still 200.
Fixes#81
fen