Alexander Brown
|
3e197b5c57
|
docs: ADR 0008 — Entity Component System (#10420)
## Summary
Architecture documentation proposing an Entity Component System for the
litegraph layer.
```mermaid
graph LR
subgraph Today["Today: Spaghetti"]
God["🍝 God Objects"]
Circ["🔄 Circular Deps"]
Mut["💥 Render Mutations"]
end
subgraph Tomorrow["Tomorrow: ECS"]
ID["🏷️ Branded IDs"]
Comp["📦 Components"]
Sys["⚙️ Systems"]
World["🌍 World"]
end
God -->|"decompose"| Comp
Circ -->|"flatten"| ID
Mut -->|"separate"| Sys
Comp --> World
ID --> World
Sys -->|"query"| World
```
## Changes
- **What**: ADR 0008 + 4 architecture docs (no code changes)
- `docs/adr/0008-entity-component-system.md` — entity taxonomy, branded
IDs, component decomposition, migration strategy
- `docs/architecture/entity-interactions.md` — as-is Mermaid diagrams of
all entity relationships
- `docs/architecture/entity-problems.md` — structural problems with
file:line evidence
- `docs/architecture/ecs-target-architecture.md` — target architecture
diagrams
- `docs/architecture/proto-ecs-stores.md` — analysis of existing Pinia
stores as proto-ECS patterns
## Review Focus
- Does the entity taxonomy (Node, Link, Subgraph, Widget, Slot, Reroute,
Group) cover all cases?
- Are the component decompositions reasonable starting points?
- Is the migration strategy (bridge layer, incremental extraction)
feasible?
- Are there entity interactions or problems we missed?
┆Issue is synchronized with this [Notion
page](https://www.notion.so/PR-10420-docs-ADR-0008-Entity-Component-System-32d6d73d365081feb048d16a5231d350)
by [Unito](https://www.unito.io)
---------
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-authored-by: GitHub Action <action@github.com>
Co-authored-by: Amp <amp@ampcode.com>
Co-authored-by: Christian Byrne <cbyrne@comfy.org>
|
2026-03-26 16:14:44 -07:00 |
|