No level editor or custom levels

Mark commented

2026-05-10 16:07:17 +00:00

Owner

Description

Only predefined levels. No user-created content.

Impact

Limited replayability.

Suggested Fix

Add a level editor with save/load.

## Description Only predefined levels. No user-created content. ## Impact Limited replayability. ## Suggested Fix Add a level editor with save/load.

Mark added the

blocked:needs-decision

label

2026-05-12 23:26:20 +00:00

Mark commented

2026-05-13 10:01:46 +00:00

Author

Owner

Feature Triage Assessment

Scope: large
Risk: low
Spec Alignment: partial
Summary: Wants a visual level editor allowing users to create, save, load, and play custom marble maze tracks to improve replayability.
Recommendation: Build as an isolated editor route (/editor) with a canvas-based grid and floating palette for placing walls, obstacles, collectibles, and spawn/finish points. Serialize to the existing LevelData interface so the engine consumes them unchanged. Persist in localStorage with JSON export/import for sharing. Add a 'Custom' tab to the track picker when user levels exist.
Rationale: Large: LevelData has 18 fields spanning 9 obstacle types, 5 power-up types, 3 collectible types, themes, hazards, ice tiles, boost pads, and checkpoints — each needs editor tooling, plus serialization, persistence, and a custom level browser. Risk is low because it's pure additive — custom levels use the existing LevelData interface, no game engine paths are touched. Spec alignment is partial since the plan.md mentions a 'level design system (tile-based or procedural)' in Stage 4 but targeted at the developer, not end-user custom content. The editor itself would be ~10-20 new/modified files.

Labeled blocked:needs-decision — unblock on Forgejo to dispatch to the fixer pipeline.

## Feature Triage Assessment **Scope:** large **Risk:** low **Spec Alignment:** partial **Summary:** Wants a visual level editor allowing users to create, save, load, and play custom marble maze tracks to improve replayability. **Recommendation:** Build as an isolated editor route (/editor) with a canvas-based grid and floating palette for placing walls, obstacles, collectibles, and spawn/finish points. Serialize to the existing LevelData interface so the engine consumes them unchanged. Persist in localStorage with JSON export/import for sharing. Add a 'Custom' tab to the track picker when user levels exist. **Rationale:** Large: LevelData has 18 fields spanning 9 obstacle types, 5 power-up types, 3 collectible types, themes, hazards, ice tiles, boost pads, and checkpoints — each needs editor tooling, plus serialization, persistence, and a custom level browser. Risk is low because it's pure additive — custom levels use the existing LevelData interface, no game engine paths are touched. Spec alignment is partial since the plan.md mentions a 'level design system (tile-based or procedural)' in Stage 4 but targeted at the developer, not end-user custom content. The editor itself would be ~10-20 new/modified files. Labeled `blocked:needs-decision` — unblock on Forgejo to dispatch to the fixer pipeline.

Mark commented

2026-05-15 13:50:36 +00:00

Author

Owner

This is broad — filed as #24 (Break down #23 into phased implementation) with 5 phases covering editor route, grid interaction, serialization, level browser, and advanced features. The decomposition issue is auto-approved for the fixer pipeline.

Mark referenced this issue

2026-05-15 13:51:04 +00:00

Break down #23 into phased implementation (5 phases) #30

Mark commented

2026-05-15 13:51:17 +00:00

Author

Owner

Decomposition filed at #30 with 5 phases: editor route, grid interaction, serialization, level browser, advanced features. This issue is addressed by the decomposition.

Mark referenced this issue

2026-05-15 13:51:18 +00:00

Break down #23 into phased implementation (5 phases) #30

Mark added

feature-request

and removed

blocked:needs-decision

labels

2026-05-15 13:51:19 +00:00

Mark added

quickfix

and removed