plans: six subsystem outline plans for the SX rewrite (store, commerce, identity, content, events, host)

Gap analysis from the five-subsystem set (acl/feed/flow/mod/search):
- store-on-sx: event-sourcing foundation the others fake with in-memory lists (build first)
- commerce-on-sx: catalog/cart/pricing/orders on miniKanren (+ store + flow)
- identity-on-sx: OAuth2/sessions/membership on Erlang (the core acl assumes)
- content-on-sx: documents/blocks/CRDT on Smalltalk
- events-on-sx: calendar/ticketing on Datalog + flow-driven delivery
- host-on-sx: the web boundary — off Quart onto native server+SXTP now, dream-on-sx next

All DRAFT outlines; substrate choices proposed, not final.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
2026-06-06 17:39:29 +00:00
parent 1e4cf25015
commit ff537bfba2
6 changed files with 521 additions and 0 deletions

82
plans/commerce-on-sx.md Normal file
View File

@@ -0,0 +1,82 @@
# commerce-on-sx: Catalog, cart, pricing & orders on miniKanren
> **DRAFT outline.** The revenue vertical. Depends on `store-on-sx` (durable
> orders) and `flow-on-sx` (checkout as a durable flow). Don't start before
> store-on-sx Phase 1 is green.
rose-ash's revenue engine — market (catalog), cart (checkout), orders (SumUp
payment, reconciliation) — has no SX subsystem. The hard part of commerce isn't
CRUD; it's **pricing**: discounts, bundles, tax, membership rates, promotions that
stack (or don't). These are relations, and a relational engine can run them in
multiple directions — forward ("what's the total?") and backward ("what promo code
yields this total?", "which line item triggered the discount?").
That's a miniKanren fit. Pricing/promotion rules are relational; cart and order
*lifecycle* (reserve → pay → fulfil → reconcile) is a durable `flow`; the order
ledger is a `store` stream. Commerce is the first real **composition** subsystem.
End-state: a catalog model, a relational pricing/promotion engine, a cart with
deterministic totals, and an order lifecycle flow with payment-webhook
reconciliation — all auditable via the event log.
## Status (rolling)
`bash lib/commerce/conformance.sh`**0/0** (not yet started)
## Ground rules
- **Scope:** only `lib/commerce/**` and `plans/commerce-on-sx.md`. May **import**
from `lib/minikanren/`, and (once they exist) `lib/store/` + `lib/flow/`. Do not
edit substrates.
- **Architecture:** prices/promotions are miniKanren relations over catalog facts;
a cart total is a *deterministic* query result (first solution under a fixed rule
order). Order lifecycle is a `flow` that suspends at the payment IO boundary.
Money is integer minor units — never floats.
- **Determinism:** promotion stacking must have explicit, tested precedence;
totals must be reproducible from the cart + catalog snapshot.
- **Commits:** one feature per commit. Progress log + tick boxes.
## Architecture sketch
```
Catalog + cart Total / order
product(id,price,tags) {:subtotal :discounts :tax :total}
│ ▲
▼ │
lib/commerce/catalog.sx lib/commerce/price.sx
— product / variant / stock facts — miniKanren pricing relations
│ — promo stacking, membership rates
▼ ▲
lib/commerce/cart.sx lib/commerce/order.sx (flow + store)
— line items, quantities — reserve→pay→fulfil→reconcile
│ — SumUp webhook = flow resume
▼ │
lib/commerce/api.sx ── (commerce/add) (commerce/total) (commerce/checkout) ──┘
```
## Phase 1 — Catalog + cart + deterministic totals
- [ ] `catalog.sx` — product/variant/stock as facts
- [ ] `cart.sx` — line items, add/remove/qty
- [ ] `price.sx` — base pricing relation, subtotal; tax
- [ ] `api.sx` + tests + scoreboard + conformance.sh
## Phase 2 — Promotions (relational)
- [ ] promo rules: percentage, fixed, bundle, member rate
- [ ] explicit stacking precedence; "best price" backward query
- [ ] tests: stacking order, mutually-exclusive promos, member vs guest
## Phase 3 — Order lifecycle (flow + store)
- [ ] order flow: reserve stock → await payment → fulfil
- [ ] payment webhook resumes the suspended flow
- [ ] order ledger as a `store` stream; idempotent reconciliation
## Phase 4 — Reconciliation + federation
- [ ] mismatch detection (paid≠ordered) as queries over the ledger
- [ ] cross-instance catalog (federated marketplace) — out-of-scope stub
- [ ] tests: webhook replay, partial refund, double-charge guard
## Progress log
(loop fills this in)
## Blockers
(loop fills this in)

82
plans/content-on-sx.md Normal file
View File

@@ -0,0 +1,82 @@
# content-on-sx: Documents, blocks & collaborative editing on Smalltalk
> **DRAFT outline.** The CMS vertical — blog, WYSIWYG editor, Ghost sync. Depends
> on `store-on-sx` (document history as an event log). Ghost/CMS sync stays a thin
> external adapter (Python/FFI) until a native replacement exists.
rose-ash's `blog` domain is content management: a block-based WYSIWYG editor,
navigation, Ghost CMS sync. A document is a tree of live blocks; editing is a
stream of operations; collaboration needs conflict-free merge. That is an object
model — blocks are objects, edits are messages, and a document is the object graph
responding to them. Smalltalk's "everything is an object responding to messages"
maps directly to a block/WYSIWYG model, and a semilattice (CRDT) merge keeps
concurrent edits conflict-free.
End-state: a Smalltalk-on-SX document model (typed blocks, structural ops),
operation log + CRDT merge for collaborative editing, versioning/history via the
event store, and a render boundary to HTML/SX. External CMS (Ghost) sync is an
injected adapter, not core.
## Status (rolling)
`bash lib/content/conformance.sh`**0/0** (not yet started)
## Ground rules
- **Scope:** only `lib/content/**` and `plans/content-on-sx.md`. May **import**
from `lib/smalltalk/`, and (once it exists) `lib/store/`. Do not edit substrates.
- **Architecture:** a document is an ordered tree of blocks (objects); an edit is a
message (`insert`/`update`/`move`/`delete`); concurrent edits merge via a
commutative (CRDT/semilattice) operation so order doesn't matter. History is the
`store` event stream; any version is a replay.
- **Determinism:** merge must be commutative + idempotent (test: apply ops in any
order / twice → same document).
- **Commits:** one feature per commit. Progress log + tick boxes.
## Architecture sketch
```
Edit op Rendered document
(insert block after id) ... HTML / SX tree
│ ▲
▼ │
lib/content/block.sx lib/content/render.sx
— typed blocks as objects — block tree → HTML/SX
— heading/text/image/embed — (reuses SX render boundary)
│ ▲
▼ │
lib/content/doc.sx lib/content/merge.sx
— ordered block tree — CRDT/semilattice op merge
— apply op, structural moves — concurrent-edit reconciliation
│ ▲
▼ │
lib/content/api.sx ── (content/edit) (content/render) (content/history) ──┐
│ │
├── op log + versions → store │
└── Ghost/CMS sync → injected external adapter (thin, non-core) ──┘
```
## Phase 1 — Block document model
- [ ] `block.sx` — typed block objects
- [ ] `doc.sx` — ordered tree, apply edit op, structural moves
- [ ] `render.sx` — block tree → HTML/SX
- [ ] `api.sx` + tests + scoreboard + conformance.sh
## Phase 2 — Op log + versioning
- [ ] edit ops as `store` events; replay to any version
- [ ] `(content/history doc)`, diff between versions
## Phase 3 — Collaborative merge (CRDT)
- [ ] commutative/idempotent op merge
- [ ] concurrent-edit tests (any order, double-apply → identical)
## Phase 4 — External sync + federation
- [ ] Ghost/CMS sync via injected adapter (import/export)
- [ ] federated documents (peer-authored blocks) — trust-gated stub
- [ ] tests: round-trip import/export, conflict on concurrent external edit
## Progress log
(loop fills this in)
## Blockers
(loop fills this in)

81
plans/events-on-sx.md Normal file
View File

@@ -0,0 +1,81 @@
# events-on-sx: Calendar, ticketing & notification delivery on Datalog
> **DRAFT outline.** The events vertical + the shared notification-delivery edge.
> Depends on `store-on-sx` (bookings ledger) and `flow-on-sx` (reminders, retrying
> delivery). Pairs with `commerce-on-sx` for paid tickets.
rose-ash's `events` domain is calendar + ticketing: recurring events, availability,
capacity, bookings. Scheduling is constraint reasoning — "is this slot free given
recurrence, capacity, and the attendee's other bookings?" — which is rule
evaluation over facts. Datalog expresses availability, recurrence expansion, and
capacity as rules; a booking is a transaction; reminders and digests are durable
`flow`s. Notification *delivery* (email/push) — needed here and by `feed/notify`
is folded in as an injected transport, extractable later.
End-state: a Datalog-on-SX events layer with recurrence expansion, availability +
capacity rules, transactional booking, and a flow-driven notification dispatcher
(reminders, digests, retries) over an injected transport.
## Status (rolling)
`bash lib/events/conformance.sh`**0/0** (not yet started)
## Ground rules
- **Scope:** only `lib/events/**` and `plans/events-on-sx.md`. May **import** from
`lib/datalog/`, and (once they exist) `lib/store/` + `lib/flow/`. Do not edit
substrates.
- **Architecture:** events/availability/capacity are Datalog facts + rules;
recurrence expands to occurrence facts within a window; a booking checks rules
then appends a `store` event (idempotent, capacity-safe). Notifications are flows
that suspend on transport IO and retry on failure.
- **Determinism:** recurrence expansion + availability must be reproducible for a
fixed window + ruleset; capacity checks must be race-safe (no overbooking).
- **Commits:** one feature per commit. Progress log + tick boxes.
## Architecture sketch
```
Event + booking Result
event(id,start,rrule,capacity) {:booked | :full | :conflict} + reminders
│ ▲
▼ │
lib/events/calendar.sx lib/events/availability.sx
— event facts, recurrence (RRULE) — free/busy + capacity rules (Datalog)
— expand occurrences in window │
│ ▲
▼ │
lib/events/booking.sx lib/events/notify.sx (flow)
— transactional, capacity-safe — reminders / digests, retry on fail
— bookings → store ledger — injected transport (email/push)
│ │
▼ ▼
lib/events/api.sx ── (events/schedule) (events/book) (events/agenda) ──────┘
```
## Phase 1 — Calendar + recurrence
- [ ] `calendar.sx` — event facts, RRULE expansion in a window
- [ ] `availability.sx` — free/busy rules
- [ ] `api.sx` + tests + scoreboard + conformance.sh
## Phase 2 — Ticketing + booking
- [ ] capacity rules; transactional booking → `store` (no overbooking)
- [ ] paid tickets compose with `commerce` order flow
- [ ] tests: capacity edge, double-book guard, conflict detection
## Phase 3 — Notification delivery (flow)
- [ ] `notify.sx` — reminder/digest flows over injected transport
- [ ] retry/backoff on transport failure (flow suspend/resume)
- [ ] tests: delivery success, retry path, idempotent re-send
- [ ] NOTE: shared with `feed/notify` — candidate for later extraction to a
`delivery-on-sx` once a second consumer is real
## Phase 4 — Federation
- [ ] cross-instance events (peer calendar) — trust-gated stub
- [ ] tests: federated agenda merge
## Progress log
(loop fills this in)
## Blockers
(loop fills this in)

100
plans/host-on-sx.md Normal file
View File

@@ -0,0 +1,100 @@
# host-on-sx: The SX web host — off Quart, onto the kernel (Dream-bound)
> **DRAFT outline.** The integration boundary that turns the subsystem libraries
> into running services, and the strangler path off Python/Quart. This is the
> dependency hub — it imports every subsystem. Decision recorded below: native
> server + SXTP **now**, `dream-on-sx` framework layer **next**, Python only at the
> external-integration edges.
The subsystems (`feed`, `search`, `acl`, `mod`, `flow`, `commerce`, `identity`,
`content`, `events`) are libraries. Something has to receive an HTTP request, route
it, call the right subsystem, and serialize the response. Today that's Python/Quart
— the one large non-SX component in the stack: separate runtime, deploy, and
failure mode. The goal is to move the web/host/domain layer onto the SX substrate
and retire Quart, **incrementally (strangler-fig), never big-bang.**
This is already underway: a native OCaml HTTP server is live in prod on
`sx.rose-ash.com` (~3ms cached, ~323 req/s, ~2MB RSS), `defhandler`/`defpage`
exist, and a partial **SXTP** protocol is specced. That is the unblocked near-term
host — no `ocaml-on-sx` dependency.
## Two layers, two timelines
1. **Now (unblocked): native server + SXTP adapter + SX handlers.** Route rose-ash
endpoints onto the SX host one at a time. Each migrated endpoint is an SX
handler dispatching to a subsystem; Quart proxies the rest until cut over.
2. **Next: `dream-on-sx` as the framework layer.** Dream gives Quart-grade
ergonomics — typed routing, middleware stacks, sessions, CSRF. It is gated on
`ocaml-on-sx` Phases 15 + minimal stdlib. **This plan is the concrete target
user that un-parks `dream-on-sx`** (see `plans/dream-on-sx.md`): "the subsystems
need an HTTP front door" is the real feature pulling Dream. Until then, do not
block migration on Dream — the native server is sufficient.
3. **Always: Python only at the edges.** External integrations — SumUp payments,
Ghost CMS, ActivityPub crypto, IPFS/Kubo — ride Python libraries today. They
stay as thin injected adapters (Python/FFI) behind subsystem interfaces until
native replacements exist. "Drop Quart" ≠ "drop every line of Python."
## Status (rolling)
`bash lib/host/conformance.sh`**0/0** (not yet started)
## Ground rules
- **Scope:** `lib/host/**` and `plans/host-on-sx.md`. May **import** every subsystem
+ the kernel's server/SXTP surface. Do **not** edit `spec/`, `hosts/`, `shared/`,
or subsystem internals — wire to their public APIs only. Host-primitive / server
changes belong in `hosts/` (out of scope) → Blockers.
- **Architecture:** a route maps (method, path) → handler; a handler is an SX fn
`request -> response` that calls subsystem APIs; middleware is composed handlers
(auth via `identity`, permission via `acl`, mute via subsystem prefs). SXTP is the
wire format between host and subsystem-as-service.
- **Migration discipline:** each endpoint moved must be behavior-equivalent to its
Quart original (golden-response test before flip). Keep a migration ledger.
- **Commits:** one feature per commit. Progress log + tick boxes.
## Architecture sketch
```
HTTP request HTTP response
│ ▲
▼ │
native OCaml http server (prod) ──────► lib/host/router.sx
(hosts/ — out of scope) — (method,path) → handler
│ ▲
▼ │
lib/host/middleware.sx lib/host/handler.sx
— auth(identity) ∘ acl ∘ mute ∘ ... — request → subsystem call → response
│ ▲
▼ │
lib/host/sxtp.sx subsystem APIs (feed/search/commerce/…)
— wire format, host↔service — called via public interfaces
└── external edges: SumUp / Ghost / AP / IPFS → injected Python/FFI adapters
```
## Phase 1 — Router + handler + one real endpoint
- [ ] `router.sx` — route table, (method,path) match
- [ ] `handler.sx` — request/response model, subsystem dispatch
- [ ] migrate ONE read endpoint (e.g. a feed timeline) end-to-end, golden test
- [ ] `conformance.sh` + scoreboard
## Phase 2 — Middleware + SXTP
- [ ] `middleware.sx` — composable auth/acl/mute/error layers
- [ ] `sxtp.sx` — host↔subsystem wire format (align with existing spec)
- [ ] migrate a write endpoint (auth + permission + action)
## Phase 3 — Strangler migration ledger
- [ ] enumerate Quart endpoints; track migrated vs proxied
- [ ] golden-response harness vs the live Quart responses
- [ ] cut over a whole domain (smallest: `likes` or `relations`) as proof
## Phase 4 — Dream framework layer (gated)
- [ ] gate: `ocaml-on-sx` Phases 15 + minimal stdlib green
- [ ] adopt `dream-on-sx` routing/middleware/session ergonomics over the same handlers
- [ ] re-home external adapters as native where replacements land
## Progress log
(loop fills this in)
## Blockers
(loop fills this in)

84
plans/identity-on-sx.md Normal file
View File

@@ -0,0 +1,84 @@
# identity-on-sx: OAuth2, sessions & membership on Erlang
> **DRAFT outline.** The identity core `acl-on-sx` assumes already exists. `acl`
> answers "may X do Y"; identity answers "who is X, and how did they prove it."
> Depends on `store-on-sx` (grant/audit ledger). Pairs with `acl-on-sx`.
rose-ash's `account` domain is the OAuth2 authorization server every other app is
a client of: silent SSO, per-app first-party cookies, grant verification,
membership. Sessions and grants are **long-lived, concurrent, individually
addressable, and expire on their own** — that is the actor model. Erlang's
processes + mailboxes map cleanly: a session is a process, token issue/refresh/
revoke are messages, expiry is a process timeout, and SSO is one process answering
many apps.
End-state: an Erlang-on-SX layer with the OAuth2 authorization-code + silent
(`prompt=none`) flows as message protocols, a session/grant registry, token
lifecycle (issue/refresh/revoke/introspect), and membership state — all auditable
through the event log, all authorization questions delegated to `acl-on-sx`.
## Status (rolling)
`bash lib/identity/conformance.sh`**0/0** (not yet started)
## Ground rules
- **Scope:** only `lib/identity/**` and `plans/identity-on-sx.md`. May **import**
from `lib/erlang/`, and (once they exist) `lib/store/` + `lib/acl/`. Do not edit
substrates.
- **Architecture:** a session/grant is a process holding its own state; the
registry routes messages by subject/client id. Tokens are opaque + introspected,
not self-validating (revocation must be real). Authorization decisions are NOT
made here — `identity` proves identity, `acl` decides permission.
- **Security:** revocation is immediate (kill the process / tombstone the grant);
no decision relies on a token that outlived its grant. Negative answers are
explicit, never "absence of a yes."
- **Commits:** one feature per commit. Progress log + tick boxes.
## Architecture sketch
```
Auth request Token / session
(authorize client scope subject) {:access :refresh :expires :grant}
│ ▲
▼ │
lib/identity/oauth.sx lib/identity/token.sx
— authz-code + prompt=none flows — issue / refresh / revoke / introspect
— as Erlang message protocols — opaque tokens, grant-backed
│ ▲
▼ │
lib/identity/session.sx lib/identity/registry.sx
— session = process, expiry=timeout — route by subject/client; SSO fan-out
│ │
▼ ▼
lib/identity/api.sx ── (identity/login) (identity/grant?) (identity/revoke) ──┐
│ │
└──────── grant + audit events → store ; permission? → acl ──────────┘
```
## Phase 1 — Sessions + tokens
- [ ] `session.sx` — session process, create/lookup/expire
- [ ] `token.sx` — issue/introspect/revoke (opaque, grant-backed)
- [ ] `registry.sx` — route by subject/client
- [ ] `api.sx` + tests + scoreboard + conformance.sh
## Phase 2 — OAuth2 flows
- [ ] authorization-code flow as a message protocol
- [ ] refresh + rotation; revocation cascades to issued tokens
- [ ] tests: full code exchange, refresh, revoke-then-use (must fail)
## Phase 3 — Silent SSO + membership
- [ ] `prompt=none` cross-app login (one session, many clients)
- [ ] membership state + per-app grant projection
- [ ] grant verification delegated cache (mirror Redis-cache pattern)
## Phase 4 — Audit + federation
- [ ] every issue/refresh/revoke is a `store` event; `(identity/audit subject)`
- [ ] federated identity (peer-asserted subject) — advisory, trust-gated stub
- [ ] tests: audit completeness, cross-instance subject mapping
## Progress log
(loop fills this in)
## Blockers
(loop fills this in)

92
plans/store-on-sx.md Normal file
View File

@@ -0,0 +1,92 @@
# store-on-sx: Event Sourcing on the SX kernel
> **DRAFT outline.** Foundation subsystem — the durable substrate the other five
> currently fake with in-memory mutable lists. Build this first.
rose-ash needs durable state: every subsystem (feed log, flow store, mod audit,
search index, acl grants) today hand-rolls an in-memory list that vanishes on
restart. They all secretly want the same thing — an append-only event log with
pure projections over it. Event sourcing makes that one substrate: the log is the
source of truth, state is a fold, durability is an IO boundary.
This is **substrate-level**, not a guest language. It lives directly on the SX
kernel's IO-suspension primitives (`perform`/`cek-resume` — the third CEK phase)
so a projection can `perform` a read/append and the kernel persists at the
boundary. Storage backends (in-memory, file, later Postgres/IPFS) are injected.
End-state: an `append`/`fold`/`project`/`snapshot` API with an injectable backend,
optimistic-concurrency on streams, replay + snapshotting, and a subscription hook
so projections (feeds, indices, audit logs) update incrementally. The other
subsystems swap their mutable list for a `store/stream`.
## Status (rolling)
`bash lib/store/conformance.sh`**0/0** (not yet started)
## Ground rules
- **Scope:** only `lib/store/**` and `plans/store-on-sx.md`. Do **not** edit
`spec/`, `hosts/`, `shared/`, or `lib/<lang>/`. May **import** the kernel's
IO-suspension surface (`perform`, the platform IO ops) — verify what's exported
before relying on it. Do not add host primitives; if a needed durable IO op is
missing, file it under Blockers (it belongs in `hosts/` / fed-prims, out of scope).
- **Architecture:** an event is `{:stream :seq :type :at :data}`. A log is an
ordered, append-only vector of events. A projection is `(fold step seed events)`.
Persistence is an injected backend `{:append :read :snapshot-read :snapshot-write}`;
the in-memory backend is the test default, real backends wire in unchanged.
- **Determinism:** replay must be pure — same log → same state, always. No clocks
or randomness inside projections; timestamps live on the event, not the fold.
- **Commits:** one feature per commit. Keep Progress log + tick boxes.
## Architecture sketch
```
Command Read model
(append stream type data) (project stream step seed)
│ ▲
▼ │
lib/store/event.sx lib/store/project.sx
— {:stream :seq :type :at :data} — fold step seed over a stream
— stream ids, seq ordering — incremental: resume from snapshot
│ ▲
▼ │
lib/store/log.sx lib/store/snapshot.sx
— append-only, optimistic concurrency — periodic fold checkpoint
— read range / read-from-seq — replay = snapshot + tail
│ ▲
▼ (perform → backend) │
lib/store/backend.sx lib/store/api.sx
— injected {:append :read ...} — (store/append ...) (store/project ...)
— mem backend (tests) | file | pg — (store/subscribe stream fn)
```
## Phase 1 — Log + in-memory backend
- [ ] `lib/store/event.sx` — event record, stream/seq helpers
- [ ] `lib/store/backend.sx` — injectable backend protocol + in-memory impl
- [ ] `lib/store/log.sx``append` (optimistic seq), `read`, `read-from`
- [ ] `lib/store/api.sx``(store/append ...)`, `(store/events stream)`
- [ ] `lib/store/tests/log.sx` + scoreboard + conformance.sh
## Phase 2 — Projections + subscriptions
- [ ] `lib/store/project.sx``(project stream step seed)`, incremental fold
- [ ] subscription hook — projection re-runs on append
- [ ] concurrency conflict surfaced as a real result, not a crash
## Phase 3 — Snapshots + replay
- [ ] `lib/store/snapshot.sx` — checkpoint a projection, replay = snapshot + tail
- [ ] compaction policy; replay determinism tests
## Phase 4 — Durable backend via kernel IO
- [ ] file/log backend driven through `perform` (IO-suspension boundary)
- [ ] crash/restart replay test (mock IO platform)
- [ ] migration notes for swapping mem → durable under a live subsystem
## Consumers (post-foundation, not in scope here)
feed/-log, flow store, mod/audit, search index, acl grant set all become
`store/stream`s. Track the migration in each subsystem's plan, not this one.
## Progress log
(loop fills this in)
## Blockers
(loop fills this in)