Files
hq/03-DESIGN/01-to-be/22-the-work-ahead.md
T
jschoubben 2e1ae061b5 The work ahead: four phases, dependency-ordered, each ending at a run
Everything decided this cycle and not yet built. Phase 0 gets the installer
green, because nothing else is testable end to end without it. Phase 1 makes the
protocol one thing and fixes the Go/TS drift the installer's own provisioning
exercises. Phase 2 stands up the private package registry ADR 0014 assumes and
publishes the SDK into it. Phase 3 adopts the substrate so one postgres and one
lavinmq serve everything, which is the hardest and needs all three above.

Order is dependency, not preference. Each phase ends at a run rather than a
paragraph, because a phase that ends at a claim is how things went missing this
cycle without anything complaining.

Claude-Session: https://claude.ai/code/session_01D6qtiYU3P9jk3pnAXyAFyx
2026-09-15 23:41:42 +02:00

4.1 KiB

layer, status, updated, decisions
layer status updated decisions
to-be in-progress 2026-09-15
02-DECISIONS/0067-genesis-is-a-pivot.md
02-DECISIONS/0074-the-wire-is-specified-not-the-types.md
02-DECISIONS/0075-two-stores-and-which-provides-what.md
02-DECISIONS/0014-no-npm-workspace.md

The work ahead

Everything decided this cycle and not yet built, in the order its dependencies allow. Each phase ends at something provable on a running mesh, because a phase that ends at a claim is a phase that went missing without anything complaining.

Phase 0 — the installer completes, once

Why first. Nothing below is testable end to end without an installer that finishes. It has been failing on regressions of mine — a stale carried builder, a diagnostic on the parsed stdout — not on the mesh.

  • 0.1 rebuild the builder image, and verify the change is in it (grep, not trust make)
  • 0.2 re-carry it, run the one-node installer, watch the phase-two base build with its own logs
  • 0.3 diagnose and fix whatever the base build actually does — the original hang, now visible
  • 0.4 all 22 checks green, twice, so it is reproducible rather than lucky

Done when. A bare machine becomes a working mesh by running the installer, and the run passes again.

Phase 1 — the protocol is one thing, and correct

Why here. The Go control plane and the TypeScript SDK disagree about what a grant carries (consumer is the module in one, the node in the other). That is exercised by the installer's own provisioning — the catalogue's database — so it belongs before more is built on it.

  • 1.1 extract the wire contracts to one specification the two implementations both conform to
  • 1.2 make Go and TypeScript agree — one meaning for consumer, the envelope's six headers emitted by both
  • 1.3 an executable conformance suite, per capability, both existing SDKs made to pass it
  • 1.4 the x-schema header written, so a body's shape can version (ADR 0074)

Done when. A fixture emitted by one implementation is read identically by the other, checked in both test suites.

Phase 2 — the private package registry, and the SDK in it

Why here. ADR 0014 says a module consumes its dependencies, the SDK included, from the private registry. Nothing installs one, so the SDK comes from a git URL (issue 053). Needs the installer (Phase 0) and gitea.

  • 2.1 gitea provides package-registry in full — the endpoints, an account the builder may publish with
  • 2.2 the builder publishes the SDK there on build, by version
  • 2.3 modules consume it by version; the git URL and the sibling-path lock are gone
  • 2.4 a second language's SDK published the same way, proving the path is not TypeScript-only

Done when. A module builds against the SDK resolved from the mesh's own registry, and issue 053 closes.

Phase 3 — nothing is special after installation

Why last. The hardest and riskiest, and it needs everything above: an installer that completes, a protocol that agrees, a registry to publish to. Issue 051.

  • 3.1 the store is adopted — raised at genesis, then held as the postgres module; one server, the control plane's records and every module's database in it
  • 3.2 the broker is adopted — one lavinmq, the / vhost for the mesh bus, a vhost per consumer that requires amqp; the second server gone
  • 3.3 an upgrade of each, proven: a store the control plane reads from, a broker over the broker, each with a stated window
  • 3.4 status can say the substrate is behind its source, which today it cannot form

Done when. A mesh built from bare metal runs one postgres and one lavinmq, and can upgrade either — so the twelve-module floor has no specialty left in it.

The through-line

Each phase leaves the mesh more able to describe and rebuild itself, and each ends at a run rather than a paragraph. The order is dependency, not preference: the installer carries everything, the protocol is what everything speaks, the registry is what everything is built from, and adoption is what makes the last two things ordinary.