Skip to content

15 — Unified Vision v3.0: The Sovereign Polyglot Ecosystem

Purpose

Declare where QCObjects is going and the transitional phases to get there — with per-repo exit criteria and migration checklists, so standard plans can be derived mechanically. This spec governs direction: when it conflicts with specs 01–14, this spec wins and the affected spec MUST be updated in the same PR.

Sources: the three roadmap drafts in roadmap/ (base vision, CosmosDB-lib structural note, consolidated update), systematized here in full as the single normative text. The drafts may be archived; this file is the record.

Vision

The transition to v3.0 is not a version bump. It formalizes QCObjects as polyglot, sovereign infrastructure: evolving from a web framework into a universal execution platform that orchestrates diverse runtimes, native-performance handlers, and cloud-native integrations — managed from a single decoupled CLI, with zero vendor lock-in and local-first AI readiness.

Why it protects current investment: handler/adapter decoupling means existing config.json shapes, $ENV()/$config() usage, and component code keep working while engines and runtimes swap underneath (proven by the CosmosDB connector pattern: same config shape, cloud-native backend).

Phase 1 — Structural Sovereignty (v3.0)

Goal: remove legal and pipeline friction for enterprise adoption.

  • License migration: full LGPLv3 → MIT across the ecosystem (mechanism: 09-license). Per-repo checklist: LICENSE.txt swap → package.json field → source headers → CHANGELOG.md → header-lint green.
  • Unified pipeline: decommission version-specific branches; single development flow with tag-based channels (latest, lts, beta) per 08-ci-conventions. Archive old tracks as tags.
  • Metadata-driven ecosystem: config.json as universal source of truth; $ENV() / $config() injection standardized across core, SDK, CLI, and every handler.
  • Rollout order (binding):
  • qcobjects-cli re-architecture first (developer entry point).
  • Core (qcobjects) + qcobjects-sdk refactor + MIT.
  • Cascade: qcobjects-commands, plugins, handlers, web components, create-qcobjects templates, eslint-config-qcobjects (+ TypeScript twin).
  • P1 exit criteria: every repo MIT-licensed with header-lint green; single-branch pipeline live with tag-channel publish proven (latest + one -beta); templates re-verified (install + serve + build) on the v3.0 line; this spec 15 + specs 08/09/10 updated to past tense.

Phase 2 — Server Abstraction Layer / High-Performance Engine (v3.1)

Goal: runtime-agnostic transport with sub-millisecond startup and massive concurrency gains.

  • Interface-driven server: decouple the CLI from Node.js's native http2 module via interchangeable Server Adapters implementing one contract: listen(config) / route(request) → backend.routes dispatch / static(request) fallback / close(). Core owns MVC routing + microservice orchestration only; transport belongs to the adapter (zero-bloat core).
  • Launch qcobjects-server-bun alongside qcobjects-server-node; engine switches by configuration flag (e.g. "serverAdapter": "bun"|"node").
  • Config service evolution: $ENV()/$config() patterns extended to complex dependency injection across all plugins.
  • P2 exit criteria: reference app serves identically on both adapters from a config-flag flip; adapter contract spec added to 02-architecture; performance delta (startup p50, concurrent connections) published in the release notes.

Phase 3 — Polyglot Execution Layer (v3.2+)

Goal: external runtimes as native microservices, orchestrated by the CLI through Runtime Bridges (all npm-installed, keyword-discovered, per spec 02).

Bridge Package Workload
Legacy validation qcobjects-handler-php Legacy monolith migration interop
Performance edge qcobjects-handler-wasm CPU-bound, crypto, secure edge (Rust/AssemblyScript/Go → Wasm)
AI & data intelligence qcobjects-handler-fastapi Persistent Python sidecars (PyTorch, LangChain, HuggingFace)
  • Sovereign AI: host local-first LLM inference + high-speed data processors on this layer (Code-Inference, WideLlama/OpenClaw-class tools, qcobjects-enterprise, collab-server for private networks).
  • Granular stream/timeout/resource control at the server level keeps stability during heavy inference or polyglot workloads.
  • P3 exit criteria: PHP + Wasm + FastAPI demo routes green on one CLI-orchestrated app; handler authoring guide published; 11-features rows flipped to stable.

Architectural invariants (all phases, normative)

  • Decoupled handlers and adapters: independent npm packages, no hard dependencies, keyword-based auto-discovery.
  • config.json remains the source of truth for runtime behavior, credentials, and routing.
  • Best tool per domain: JS/TS for web, Rust/Wasm for speed, Python for data intelligence — one unified deployment pipeline.

End of Life (v2.x, normative)

Line State Support
v2.3.x, v2.4.x-beta, v2.4.x-ts, v2.5.x-beta lines (dotted versions, e.g. v2.3.1, v2.4.1-beta) EOL, archive tags only (CLI also carries archive/v2.4-beta, archive/v2.4-ts prefixed tags) No patches, features, or community support
v3.0 track Current Full maintenance
Legacy enterprise SLA only Custom patching / guided migration via consulting contracts

Migration to the v3.0 track is strongly recommended (MIT + unified pipeline + adapters). Open-source sustainability: Open Collective / GitHub Sponsors.

Planning standard (normative)

Future plans MUST cite the phase (P1/P2/P3), list touched specs, state acceptance per those specs' Verification sections, and name the exit criteria above they advance. A plan that changes direction MUST update this spec first.

Verification

  • P1 exit: all criteria in Phase 1 list green.
  • P2 exit: Bun adapter serves the reference app from a config flag flip.
  • P3 exit: PHP + Wasm + FastAPI demo routes all green on one CLI-orchestrated app.