{"repo":"trickle-labs/pg-trickle","free":true,"listed":false,"github":"https://github.com/trickle-labs/pg-trickle","clone":"git clone https://github.com/trickle-labs/pg-trickle.git","description":"A PostgreSQL 18+ extension for streaming tables with incremental view maintenance, powered by differential dataflow in Rust.","language":"Rust","stars":133,"topics":["dvm","extension","ivm","materialized-views","postgresql","streaming"],"license":"Apache-2.0","category":"databases-storage","readme_excerpt":"pg trickle Pre-1.0 Notice: pg trickle is under active development and has not yet reached a stable 1.0 release. The core feature set is extensive and thoroughly tested, but APIs and configuration options may still change between minor versions. See ROADMAP.md for the planned path to 1.0. For a plain-language description of the problem pg trickle solves, the differential dataflow approach, and the hybrid CDC architecture, read ESSENCE.md . Stream Tables for PostgreSQL 18 pg trickle brings declarative, automatically-refreshing materialized views to PostgreSQL, inspired by the DBSP differential dataflow framework (comparison). Define a SQL query and a schedule; the extension handles the rest. A stream table is a table defined by a SQL query that stays up to date automatically as the underlying data changes. You write to your base tables normally — INSERT , UPDATE , DELETE — and pg trickle propagates the changes downstream. No polling, no manual refresh calls, no application-level orchestration. The key difference from REFRESH MATERIALIZED VIEW is that pg trickle uses Incremental View Maintenance (IVM) : when a row changes, only the effect of that row on the query result is computed. If you insert one row into a ten-million-row table, pg trickle processes one row — not ten million. At high change rates it falls back to a full recompute automatically; at low change rates (the common case) the speedup is substantial. The TPC-H benchmarks show 5–90× measured improvements across the ","default_branch":null,"files":null,"tree":[],"storefront":"/r/trickle-labs","claimed":false,"request_supported":{"post":"https://gitbuyer.com/r/trickle-labs/pg-trickle/request-supported","requests":0},"note":"indexed from public GitHub; nothing is for sale on this page. Clone it from GitHub. Paid listings live at /search."}