Pillar

Appearance

Why Pillar ​

Pillar helps you build rich domain models and event‑sourced systems in Laravel — without forcing a rigid, heavyweight project layout.

Pragmatic DDD, not dogma ​

  • Pillar supports classic DDD patterns (aggregates, repositories, commands/queries, upcasters, snapshotting), but it doesn’t force a complex package/module structure.
  • Start from your current codebase and adopt the parts you need. Keep controllers + Eloquent where they make sense; introduce aggregates where the domain benefits.

Fits non‑DDD apps, too ​

Use Pillar just for auditing or for event‑driven islands inside a conventional app:

  • Model a few event‑sourced aggregates to capture important business facts.
  • Keep the rest state‑based with a custom repository — both styles live side‑by‑side.
  • Use projectors to maintain read models and denormalized views for queries and dashboards.

Performance by design ​

  • Generator‑based streams in the event store enable true streaming of large histories.
  • Pluggable fetch strategies (load‑all, chunked, streaming) so you can tune per aggregate.
  • Snapshotting policies (Always / Cadence / On‑Demand) avoid long replays.
  • Optimistic concurrency handled by the session; no extra round‑trips.
  • Reflection metadata caching in the default serializer for fast (de)serialization.

Solid feature set ​

  • 🧠 Aggregate sessions (Unit of Work) — find(), mutate, commit()
  • 🗃️ Pluggable event store with generator‑based streams & optimistic locking
  • 🧵 Fetch strategies (load‑all / chunked / streaming)
  • 🧬 Versioned events & upcasters
  • 💾 Snapshotting policies (Always / Cadence / On‑Demand)
  • 🧩 Object serialization — JSON by default, MessagePack built-in, or custom serializer
  • 🔒 Payload encryption — pluggable cipher, per‑event overrides
  • 🖥️ Event stream browser Web UI — browse streams and timelines and inspect payloads
  • ⏱️ Point‑in‑time reads — load up to aggregate/global sequence or date via EventWindow
  • 🎭 Aliases for readable event names
  • 🔁 Safe replays to rebuild projections
  • ⚡ CQRS — projectors and query bus for a fast, scalable read side
  • 🧰 Facade + buses for quick wiring
  • 🛠️ Pillar Make: Bounded Context/Command/Query Scaffolding

Use‑cases ​

  • Systems that need an audit trail of business decisions and changes.
  • Collaborative domains with long‑lived aggregates (documents, orders, accounts).
  • Integrations that prefer events as the source of truth and read models for queries.
  • Legacy apps where you want to introduce event sourcing gradually in critical areas.

Next steps ​

  • Start with the Getting started guide.
  • Then read the core concepts in this order: Aggregate roots → Aggregate IDs → Aggregate sessions.
  • Explore the Event Store and Snapshotting when you outgrow simple cases.