The it, pt and uk catalogs merged from main still carried the two Snow
button strings, which the catalog tests report as stale. They are
pruned with npm run i18n -- --prune.
A Scene Director shot saved with Snow now normalizes to Normal when the
project loads, so the shot list shows NORMAL rather than an untranslated
SNOW for a shot that plays on Normal. The README's style keys read 1-6.
createView threw on style 'snow', so an embed host or a
show_in_gods_eye_view call holding a view from before Snow was retired
was refused outright, camera and layers included. viewFromParams read
style=snow as the recipient's own style, while the app opened the same
link on Normal.
Both now read a retired style as Normal; truly unknown names still
throw. The retired list moves to a browser-free module shared by the
view API, the share codec and visual settings.
Snow is no longer offered anywhere a style can be chosen: the Visual
Presets tray, keyboard shortcut 7, the Cockpit vision carousel, and the
set_visual_style voice action. The shader and its labels go with it;
weather snowfall and the SNOW weather condition are unaffected.
Share links keep working. Snow's enum code (6) stays reserved, so a
compact or long-form link that names it opens on Normal and drops the
Snow shader parameters without reporting them as unknown. A Scene
Director shot saved with Snow plays on Normal too. Snow's frozen share
defaults matched Normal's apart from those parameters, so no other
setting in an old link changes meaning.
The legacy voice-schema pins restore Snow before hashing, so the
unchanged pins still prove nothing else moved; the complete Realtime
payload digest is re-derived.
Mirror the dock trays and their pin, the compact Radio window and the
phone top-bar actions in right-to-left; keep the Cyber cockpit's Data
Layers stack on the left edge. Isolate the Space Missions counter,
landing coordinates and CCTV summary parts, format their numbers with
the locale's numerals, decide pack shot titles by the marker binding,
and place Bhote Koshi media cards between the panels on either side.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Localize a bootstrap shot title without a source pack only when the
Nepal pack marker binds that title to the shot's id, so an operator
shot renamed to a bootstrap title in the built-in scene stays as typed.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Localize a legacy Nepal base shot only when the applied pack marker binds
that title to this exact shot id, so a user shot renamed to a built-in
title stays as typed in every locale.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
A pack marker with shotBindings now vouches only for the shot bound to a
legacy bootstrap title, so a user shot typed with the same title keeps its
text instead of showing the localized built-in name.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Localize a pack shot title only for the shot the pack marker binds to that
title, so an operator's shot named like a pack shot shows as typed. Markers
saved without shotBindings keep the title-only check.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Japanese has no letter case, so the HUD summary keeps units as the
catalog writes them (13 km), matching the cockpit and the ja HUD CSS.
The ja '{value} KM', '{value} KM/H' and '{value} KM/S' strings are now
lower-case like zh-CN.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Conflicts in es, ru, zh-CN and zh-Hant catalogs: both sides appended keys
at the end (cyclone errors from this branch, routing refusals from main);
kept both.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The HUD builds place context for its AI summary whenever the camera
settles, and each build paid for up to four reverse geocodes plus a Nearby
Places search, even when no OpenAI key was configured to use them.
- Ask GET /api/openai/hud-summary/status once; with no summary configured,
or after an unconfigured reply, make no place lookups for the page.
- Reverse-geocode at most two points: the view center and one viewport
point.
- Reuse a lookup until the view center moves a quarter of the visible
width, the camera's distance to it doubles or halves, or the camera
crosses a lookup height. The reused points are asked again, so only a
timeout, failure or empty answer spends.
- Repaint the last AI reply when a small pan leaves the context unchanged.
Cached-only hosts and voice context lookups are unchanged.
