• 4 Posts
  • 253 Comments
Joined 2 years ago
cake
Cake day: July 16th, 2023

help-circle

  • NotSteve_@lemmy.catoLinux@lemmy.mlFan of Flatpaks ...or Not?
    link
    fedilink
    arrow-up
    14
    arrow-down
    3
    ·
    edit-2
    6 days ago

    For sure and I agree that should be enough but the average person is not good with computers and they don’t want to learn. They won’t understand the nuances of different distributions of Linux. Like try explaining the difference between a .deb, a .tar.gz, and a .rpm to a person who’s already hésitent about using Linux. Flatpak solves that by just having one download that any Linux install can use


  • NotSteve_@lemmy.catoLinux@lemmy.mlFan of Flatpaks ...or Not?
    link
    fedilink
    arrow-up
    10
    arrow-down
    5
    ·
    edit-2
    6 days ago

    Oh 100% but have you tried to explain how to use one to a computer novice? Like yes, the answer is usually “they should just…” but novice users will never. With flatpak, they get an experience similar to how MacOS works and a bit like how .exes work and it Just Works™️

    Edit: like I’ve had trouble showing people how to use the GNOME App Store which could not be any more simple. Anyone who has been convinced to install Linux already feels way out of their element so making everything feel as natural as possible is essential (and I mean, flatpaks are awesome anyway)


  • NotSteve_@lemmy.catoLinux@lemmy.mlFan of Flatpaks ...or Not?
    link
    fedilink
    arrow-up
    49
    arrow-down
    4
    ·
    edit-2
    6 days ago

    I love installing things from the CLI and prefer to only do it that way but Linux needs a single click install method for applications if it’s ever going to become a mainstream OS. The average person just wants to Google a program, hit download and install. If not that then they want to use a mobile-like App Store.

    Flatpak is kind of perfect at achieving both those things












  • I guess it’s less the response itself but the response objects in the codebase. The main code base I work on is fully typed Python/FastAPI but we depend on an API from another team that was, up until recently, untyped entirely. The responses were just dictionaries that were dynamically generated randomly in 10 layer deep functions. There was absolutely no way to tell what they would actually return, let alone what types they would be. It’s like, if you see a function that’s just called get_big_fat_entity_findings() with no typing, how would you handle that without spending an hour reading through the code line by line? Typing is important for any project that’s bigger than a quick script