I submitted PodcastSaver (https://podcastsaver.com) before but the reason it's interesting now is that I've started converting it into a live search engine test-bench.

I've discussed it a bit here[0], and this idea has been kicking around in my head for a while so I got a chance to do some related writing about it with Supabase[1].

The basic idea is to use a modest piece of the podcast index[2] as a place to test out different new age search engines against each other.

So far there are two engines running:

- Postgres FTS + pg_trgm (tuned -- indices are there, I did some EXPLAINing earlier today to tighten things up, but still all built-in tech)

- Meilisearch (untuned -- just stand it up, give it resources and put in documents)

To that effect, I've added a "nerds" page you should peruse: https://podcastsaver.com/nerds

On that page you can:

- choose your search engine

- choose whether to force disable the cache (obviously... you'd want that, for the results to mean anything, but for regular people surfing the cache is on!)

As far as actually getting the podcast search really good, there is a ton of curation left to do so it's a subpar consumer product still, but it's interesting from at least this angle! Going to add more search engines later but who knows when (this project was supposed to be short!).

I can't add every engine on the huge list of new-age search engines[3], but I can say that I will get to highlighting all of them in Awesome F/OSS[4]... Eventually.

[0]: https://news.ycombinator.com/item?id=33316029

[1]: https://supabase.com/blog/postgres-full-text-search-vs-the-rest

[2]: https://podcastindex.org/

[3]: https://news.ycombinator.com/item?id=33317232

[4]: https://awsmfoss.com