☆ Yσɠƚԋσʂ ☆

  • 39 Posts
  • 18 Comments
Joined 5 years ago
cake
Cake day: January 18th, 2020

help-circle
























  • It’s great to hear things mostly worked out. Stuff like scaling bottlenecks is definitely tricky to catch until you have serious loads on the system, but sounds like the fixes very mostly trivial validating overall design. It also looks like you managed to get a way with a fairly simple stack by leveraging Postgres and Rust. I’ve had really good experience with using pg myself, and really don’t see a point in using anything else now. You can use it both as a relation db and a document store, so it’s extremely flexible on top of being highly performant. Keeping the stack simple tends to be underappreciated, and projects often just keep adding moving pieces which end up adding a lot of overhead and complexity in the end.







  • ☆ Yσɠƚԋσʂ ☆@lemmy.mltoAnnouncements@lemmy.mlLemmy AMA March 2025
    link
    fedilink
    English
    arrow-up
    31
    arrow-down
    1
    ·
    1 month ago

    What have been the biggest challenges with the project over the years, both in terms of technical and non technical aspects. I’d be interesting to hear a bit of retrospective on how has the stack’s been working out, and what surprises you might’ve run into in terms of scaling and federation. What recommendations you’d make based on that and what you would’ve done differently knowing what you know now.