The future of the extension ecosystem
summary
History and numbers frame the case PGXN holds 396 extensions while a GitHub gist finds ~1,200 and ecosystems like crates.io have 146,000 distributions, so the talk argues the current scatter harms discovery and uptake. Concrete work described includes a new PGX V2 project at Tempo, a command line client for scaffolding, CI and release feeds, binary packaging examples from trunk and PGX man, and an interaction system with webhooks and release streams plus canonical registry, metadata and curation to enable fast binary installs and automated packagers.
AI-written summary · about
chapters (12)
- 0:00Introduction and speaker background
- 0:48History of Postgres extensions
- 6:38PGXN's origins and evolution
- 10:30Challenges in extension ecosystem
- 16:14Current state and adoption issues
- 18:20PGX V2 and community engagement
- 20:05Needs of extension authors
- 29:25Needs of packagers and administrators
- 40:29Needs of end users
- 45:29Industry's role in extensions
- 49:01Vision for future architecture
- 51:33Q&A session
related talks
49mSupporting extensions, but really now
16mRecent Activity, Trends, and Priorities at Open VSX
27mExtensions Are Dead, Long Live Extensions!
16mWho are 'the ggplot2 extenders' and how to become one
42mThe Good, the Bad and NPX
39mThe JupyterLab Extension Ecosystem- Trends from PyPI +GitHub
09mDocker Extensions: The Latest and Greatest
45mNext Gen Package Management