Bye, Google Search

6 months ago

The blog previously used Google for search but switched to Pagefind due to Google's inability to reliably find pages.
Pagefind is a static search library that works efficiently with static file sites, requiring minimal setup.
The indexing process for Pagefind took about a minute for 2.5 million words on a VPS, raising questions about scalability for larger datasets.
Pagefind's index is client-side, fetching necessary pieces via HTTP GETs, making it fully static.
Customization of Pagefind's UI involved some CSS tweaking and adding specific data attributes for metadata.
The blog contains unpublished content, requiring manual cleanup to prevent it from appearing in search results.
Pagefind supports advanced search features like matching across verb forms and singular/plural, with exact match options using quotes.
The author appreciates Pagefind's multilingual capabilities, though they primarily write in English.
A cron job is set up to re-run the Pagefind indexer daily for updated search results.
The transition to Pagefind was smooth, with the author expressing satisfaction with its performance and features.

Hasty Briefsbeta