search

AfterClass-io / afterclass.io-v2

[alpha] https://new.afterclass.io
https://afterclass-io-v2-git-main-afterclass.vercel.app/
MIT License
7 stars 2 forks source link

[Research] explore feasible options for search functionality #72

Open davidlhw opened 9 months ago

davidlhw commented 9 months ago

Crucial Decision Factors (ranked by priority):

  1. User experience
  2. Transparent data management policies
  3. Self-hostability
  4. Cost (should not cost >50 USD/mth, ideally should be free / ~10 USD/mth)

Consider options like:

Also consider (with a pinch of salt) comparison tables from the various search solutions:

onebignick commented 5 months ago

I will be taking this up, can assign to me?

davidlhw commented 1 day ago

consider using lightweight client-sided in-memory search options like

where the process would look something like:

  1. fetch data from postgres => form a datastructure that is the minimum required to render the /search?q=<query> page (ie. prof id, prof slug, review count etc)
  2. load client-sided search script
  3. index all results from 1a
  4. search-as-you-type experience, from querying the client's in-memory mini search engine