OSS projects tracked in the garden. Each note is a defuddle capture of a
project’s README, enriched with language/license/stars from the
GitHub API (scripts/oss-enrich.py).
Database view
The full table — sortable, with stars — lives in OSS.base.
Projects
Search & NLP (see Search solutions for static sites, Polish stemming and lemmatization):
- zbsearch — Orama fork by the original team; preferred for our EN+PL work
- pagefind — static search, chunked WASM index, PL/CS stemming
- meilisearch — server-side, pol/ces/slk locales, hybrid search
- typesense — server-side, leanest RAM (~30 MB idle)
- flexsearch — the engine inside Quartz search today
- lunr.js — classic SSG search; lunr-languages adds PL
- orama — upstream of zbsearch (sk=Sanskrit gotcha)
- qmd — local hybrid search for the wiki/agent side
- snowball — stemming DSL; official PL+CS algorithms
- pystempel — Python Stempel port (Lucene’s PL stemmer)
- python-morfeusz — Morfeusz 2 bindings (PL lemmatization)
Garden & tooling: