Repository navigation
id: research: prior art — capability systems, OCAP, Spritely Goblins/OCapN, CRDT shared worlds #5989
Description
Part of the worlds design program: #5955. Statements S77 and S110 of the 2026-11-04 brief: the last part of the brief is research into capabilities, state of the art, predecessors, and prior art.
Scope (Drew, in substance)
A strong influence would definitely be Spritely Goblins and all the variety of research and innovation around CRDTs, OCapN, and shared virtual worlds and real estate. Survey that space and map it against our model:
- Object-capability systems: E, Joule, Cap'n Proto, seL4-style capability hardware/OS work — the theory our "pass the platform in, nothing ambient" model (programming-model child of #5955) descends from. Where do we follow ocap discipline exactly, and where do we deliberately deviate (e.g., compile-time platform ceilings are not classic ocap)?
- Spritely Goblins / OCapN: transactional actors, distributed object protocols, petnames — mapped against our identity/invite system (#5751) and writer-set authority (#5545).
- CRDTs and shared state: which of our world-state problems are CRDT-shaped (presence, shared documents, collaborative building) and which are authority-shaped (game state under a world author's rules) — the line matters and the survey should draw it.
- Shared virtual worlds & real estate: the land/ownership line from Habitat through Second Life parcels to modern attempts — what "real estate" means when the world is user-hosted (self-hosting child of #5955).
- Adjacent modern work: WebAssembly component model / WASI capabilities (the cross-cutting layer of the capability-provider survey #5931 — coordinate, don't duplicate: that survey covers provider classes; this one covers capability theory and protocols).
Done looks like
A research doc with a per-system entry and an explicit "adopt / adapt / reject, because" line for our design for each — feeding the programming-model, capability-model, and self-hosting children.
Metadata
Metadata
Assignees
Labels
Package/app: id (pkgs/ or apps/ in this repo)A design proposal (design-architect): multi-step, phased, decision-bearingOrdinary priority (the default)Queue task is parked (waiting on CI, time, or an event)Deliberately paused with a recorded resume conditionQueue packet kind: verify, reproduce, and route this itemQueue worker lifecycle state
Activity
developing-today commentedon Nov 5, 2026
CODE-1655
dezren39 commentedon Nov 5, 2026
Wave-2 tie-in (finding F5) — part of this survey was already paid for once.
On 2026-10-10, a dedicated research task (OpenCode session
ses_edba8ff24ffemYX3iqlHI2vktt) investigated: the OCapN spec + Rust implementation state, Keyhive (threat model, ciphersuite, crates.io license/maturity), p2panda auth/core/sync (versions, licences, iroh use), Spritely Goblins concepts to borrow (Racket — concepts, not importable), and Petnames (concept-only?). License verification was a required part of that task. The Goblins vat/actor and Petname findings are inputs to the programming-model (#5963) and permission-UX designs — reuse them rather than re-surveying.Record:
workspace/scratch/worlds-capabilities-brief-2026-11-04/conversation-tie-ins.md(F5).dezren39 commentedon Nov 5, 2026
Wave-2 follow-ups filed from the prior-art research (Lane B). Granular sub-scopes of this issue, each carrying the relevant research findings + citations:
Research source:
workspace/scratch/worlds-capabilities-brief-2026-11-04/prior-art-research.md. Full wave-2 mapping:workspace/scratch/worlds-capabilities-brief-2026-11-04/wave2-map.md.dezren39 commentedon Nov 5, 2026
🤖 Agent note · handed off
106 remaining items
dezren39 commentedon Nov 6, 2026
🤖 Agent note · handed off
dezren39 commentedon Nov 6, 2026
🤖 Agent note · handed off