feat: stage simulation garden first read
This commit is contained in:
@@ -29,15 +29,18 @@ Terrain3D world is running in `main.tscn`, NPCs are alive in the terrain-backed
|
||||
village, the selected-NPC reason inspector is fed by real task scoring, and the
|
||||
cinematic/debug toggle produces a shareable view without changing simulation
|
||||
state. A follow-up pass adds small task glyphs above active NPCs, so the
|
||||
cinematic view now preserves task intent after the debug labels are hidden.
|
||||
cinematic view now preserves task intent after the debug labels are hidden. A
|
||||
camera-staging pass then gives the runtime capture a wider village view and adds
|
||||
authored dirt path strips that make the task loop readable from the first frame.
|
||||
|
||||
The frame also makes the next presentation risks plain:
|
||||
|
||||
- the camera still favors verification over composition;
|
||||
- placeholder characters, houses, and work props remain useful but visibly
|
||||
prototype-grade;
|
||||
- terrain texture variation reads better than the old flat map, but paths,
|
||||
landmarks, water, and foreground silhouettes need stronger first-read staging;
|
||||
- terrain texture variation and path strips read better than the old flat map,
|
||||
but landmarks, water, and foreground silhouettes need stronger first-read
|
||||
staging;
|
||||
- task glyphs help, but the cinematic view still needs clearer work-site props
|
||||
and route composition so intent is readable before the viewer learns the icon
|
||||
language.
|
||||
|
||||
Reference in New Issue
Block a user