For the record I'm not a weeb. I scraped some twitter data of people posting comifuro catalogue and i made a frontend for that data
- rewrite this shit lmao
- just use sqlite even locally
- think about deduplication
- doing dailies was a bad idea. Just do live and use tweets as checkpoints
- handle errors better and do retries
- strip the hashtags when storing
- try other search methods (vector, fuzzy, etc)
- search goodies. ignore tweets, ignore users, sort by user, sort random, filter by day, etc
- image carousel
- infinite scroll
- virtualize
- pulling metadata from r2 was stupid. just do an actual db and cache routes