Development Roadmap & Versioning

Project: foundry
Organization: foundkit

• Pending
• [/] In Progress
• Completed

• 1.1 Core Platform & Application Decoupling

  • Create facade crate foundry with FoundryApp::builder() and unified prelude.
  • Full separation between framework runtime and independent user applications.
  • Package and distribute via standard Cargo dependencies (foundry = "0.1").

• 1.2 Subsystem Standard & Dynamic Routing

  • SubsystemModule trait for registering business modules, custom routes, and admin pages.
  • Static and dynamic external subsystem discovery engine.
  • Topic-scoped RBAC and Context extraction.

• 1.3 Storage & Admin Control Plane

  • PostgreSQL Zero-DDL dynamic model schema engine.
  • Auto-generated RESTful CRUD endpoints (/api/v1/s/{system_slug}/{model_slug}).
  • Embedded Admin SPA Shell with iframe sandbox bridge (FoundryBridge).

• 1.4 Developer CLI Tooling (foundry-cli)

  • foundry new <project>: Scaffold standalone user application.
  • foundry system new <slug>: Scaffold business subsystem.
  • foundry migrate: Apply baseline migrations.
  • foundry validate / system list: Project specification validation and subsystem discovery tooling.

• 1.5 Reference Example & Automated Testing

  • examples/blog_platform: Full-featured standalone reference application.
  • End-to-end integration and CLI smoke tests in CI.

• [/] 2.1 Sandboxed WebAssembly (Wasmtime) Plugin Runtime

  • Hot-reloadable WASM subsystem modules.
  • Safe sandbox boundary for multi-tenant extensions.

• [/] 2.2 OpenAPI 3.0 & Swagger UI Integration

  • Auto-generate OpenAPI schemas for dynamically created data models and subsystem routes.
  • Embedded Swagger UI explorer in Admin Shell.

• 2.3 Multi-Database Support & Pluggable Storage SPI (PostgreSQL, MySQL, MariaDB, MongoDB)

  • Complete architecture design & implementation roadmap: Storage Engines & Multi-Database Architecture.
  • Storage Engine SPI & Smart Detection: Abstract StorageDriverProvider and StorageRegistry, supporting protocol sniffing from DATABASE_URL and explicit selection via DATABASE_TYPE.
  • Method-Based Decoupling: Encapsulate raw pools completely within the Database facade (db.records(), etc.).
  • Phased Implementation Matrix & Progress:
    • Phase 0 (1st Integration - [x] Completed): Core SPI & Trait abstraction; refactor PostgreSQL into PostgresProvider with JSONB & GIN inverted index.
    • Phase 1 (2nd Integration - [x] Completed): MySQL 8.0+ / 8.4+ strategy with auto-increment, DATETIME(6), deferred join pagination, and transaction atomic re-query.
    • Phase 2 (3rd Integration - [x] Completed): MariaDB 10.5+ dialect and driver adaptation, dedicated migration schema, and covering composite indexes.
    • Phase 3 (4th Integration - [x] Completed): MongoDB 6.0+ native document strategy with BSON mapping, atomic _counters sequence compatibility for i64 primary keys, and Wildcard indexing (data.$**).
    • Phase 4 (Production-Readiness & Release - [x] Completed): Upgrade foundry migrate for dialect-aware migrations, containerized test suites, and developer guides. (Oracle DB removed from scope due to heavy runtime dependencies).

• 3.1 Cluster Orchestration & Distributed Pub/Sub

  • Real-time WebSocket event streaming.
  • Multi-node cache synchronization via Redis Pub/Sub.

• 3.2 Dynamic Webhook Dispatcher

  • Configurable webhook endpoints triggered on entity mutations.
  • Retry queue with exponential backoff.

Foundry follows Semantic Versioning (SemVer):

• Major versions (1.x.x): Breaking changes to public APIs with automated migration guides.
• Minor versions (0.x.x, 1.x.0): New feature additions, builder extensions, and backward-compatible trait defaults.
• Patch versions (0.1.x): Backward-compatible bug fixes and internal performance enhancements.

Upgrading in a user application:

cargo update -p foundry