A Developer Envisions Treating the Open Web Like a Queryable Database
A developer working in the AT Protocol ecosystem (the tech behind Bluesky) published a piece imagining what it would look like if blog posts across the internet were structured data you could query directly, rather than unstructured HTML that has to be scraped or indexed by search engines.
The idea builds on protocols like AT Protocol and other decentralized data schemas, where content is published with consistent, machine-readable structure. Instead of every blog living in its own bespoke format, posts could carry standardized fields for authorship, timestamps, tags, and content, letting any client run something conceptually like 'SELECT * FROM internet.blogposts.'
The post drew a modest but engaged discussion on Hacker News, with commenters weighing in on feasibility, existing prior art like RSS and ActivityPub, and the tradeoffs of centralized indexes versus federated schemas.