Nov 13, 2026, 9:07 PM@xgranade I'm trying to be very careful about the data format I make here because I want as few migrations as necessary and possible, and that if I DO have to do them they are simple and yeah, can be done manually or automatically, both server and desktop side.
I doubt I am going to get away with version 0/1 of the database schema being the final one but I doing my best to get as close to that as I possibly can.
Nov 13, 2026, 9:10 PM@xgranade But yeah, definitely the idea here is that I cannot, and never will, be able to blow away a production database, and so the less one has to fuck with it during upgrades, the better. This is also why my first round of real, actual testing will be done against something I am hosting, so that as I test it against real data I'm only fucking myself over by scraping my own service a trillion times...
Somewhere in the back of my mind, I'm wondering how easy it would be to serve up a wikipedia copy via HTTP and let myself scrape against it. Worst case scenario, I just hit the hell out of my own social media, I guess, because I imagine once I see how the database grows and performs under load I might have some very different ideas about what the database looks like...