Add semantic positioning system for scene layout #1

Closed
opened 2026-02-20 20:20:12 +00:00 by Mark · 1 comment
Owner

Problem

Currently, scene layout requires hard-coded normalized coordinates (0-1 range). This is unintuitive and requires trial-and-error.

Solution: Scene Anchors

  1. Scene-level anchors - Define named reference points:

    "anchors": { "table_surface": [0.5, 0.65] }
    
  2. Anchor references in transforms:

    "transform": { "anchor": "table_surface", "anchor_offset": [-0.12, 0.15] }
    

Implementation Status

  • Schema + Renderer (commit 352a3a8)
  • Prefab/actor defaults should also support anchors
  • Camera could use anchors for framing

Why This Matters

  • Semantic, not numeric — "Put character at table" vs "Y=0.55"
  • Scene-relative — Anchors defined per-scene
## Problem Currently, scene layout requires hard-coded normalized coordinates (0-1 range). This is unintuitive and requires trial-and-error. ## Solution: Scene Anchors 1. **Scene-level anchors** - Define named reference points: ```json "anchors": { "table_surface": [0.5, 0.65] } ``` 2. **Anchor references in transforms**: ```json "transform": { "anchor": "table_surface", "anchor_offset": [-0.12, 0.15] } ``` ## Implementation Status - [x] Schema + Renderer (commit 352a3a8) - [ ] Prefab/actor defaults should also support anchors - [ ] Camera could use anchors for framing ## Why This Matters - **Semantic, not numeric** — "Put character at table" vs "Y=0.55" - **Scene-relative** — Anchors defined per-scene
Author
Owner

Feature request (semantic anchor positioning system). Schema + renderer already implemented per commit 352a3a8. Remaining work appears to be content/authoring, not code. Tabling.

Feature request (semantic anchor positioning system). Schema + renderer already implemented per commit 352a3a8. Remaining work appears to be content/authoring, not code. Tabling.
Mark 2026-05-28 03:36:10 +00:00
Sign in to join this conversation.
No description provided.