selfoss

Selfoss packages feed aggregation into a twenty-five megabyte PHP application that runs on SQLite, MySQL, or PostgreSQL without runtime daemons. Created in 2012 by Tobias Zeising and maintained by Jan Tojnar, it collects RSS and Atom subscriptions alongside custom scraped feeds into a single timeline. Homelab operators deploy it on standard LAMP or container stacks to retain complete custody over reading history.

Why Selfoss instead of Netvibes

Netvibes centralizes feed subscriptions and reading telemetry on cloud infrastructure owned by Dassault Systèmes. Every feed request, bookmark, and article read generates analytics on corporate servers, tying content consumption to hosted accounts and commercial tiers. When publishers block cloud crawlers or change access terms, hosted readers offer little recourse. Selfoss shifts the fetching and storage pipeline to private hardware, ensuring that subscriptions and archived articles never touch an external vendor.

The operational difference reflects contrasting goals. Netvibes provides a hosted multi-tab dashboard with business analytics, financial monitoring, and app widgets, eliminating server maintenance. Selfoss deliberately focuses on feed and data-stream ingestion without widget clutter. Feeds poll via system cron using cliupdate.php, with article extraction handled locally by embedded Graby and FullTextRSS rules. Homelab administrators trade managed hosting for complete data sovereignty, taking responsibility for PHP updates and directory backups.

What it does well

• Zero-daemon runtime: operates as a standard PHP 8.2 application on Apache or Nginx, keeping idle memory under thirty megabytes on modest homelab servers.
• Flexible database backends: supports zero-configuration SQLite for simple setups alongside MySQL, MariaDB, and PostgreSQL for dedicated database hosts.
• Local full-text extraction: uses embedded Graby and FullTextRSS patterns to scrape full article bodies from truncated feeds directly on the server.
• Cron-driven polling: delegates feed updates to system crons or systemd timers via cliupdate.php, avoiding unpredictable background web workers.
• Extensible spout architecture: custom PHP connectors pull updates from non-standard websites and REST endpoints into the main timeline.
• Single-directory backup: keeps configuration, caches, icons, and SQLite data within the directory for painless snapshot backups.