• 0 Posts
  • 11 Comments
Joined 3 years ago
cake
Cake day: June 15th, 2023

help-circle
  • When Purism unveiled their first linux phone, I was ecstatic. Finally, someone takes us nerds seriously! They were going to compatible with the existing linux ecosystem, they were going to use established linux phone tech, and finish what these projects started. I was about to impulse preorder it, but then I decided to wait instead.

    The rot started immediately.

    First, they decided against using kde mobile, which was already in heavy development at the time. They were instead going to use gnome. Which did not have a mobile version yet. The reason they gave was for “convergence” with their other products. Some time after that the gnome leadership disagreed with their design direction and started work on the actual gnome mobile. So purism started reinventing the wheel, developing another linux mobile UX from scratch.

    A bit later, a conference was organised by older linux phone projects. It was supposed to be about what not to do, why all the other failed, and perhaps how to build commons that all linux phone projects could rely on. Purism declined to participate, and the conference plans fell apart.

    Purism then launched their own appstore, incompatible with all the other appstores from the other projects.

    When the phone finally released, it was ancient hardware at an insane price. After that, purism raised the price further, in order to chase their new dream, 100% US manufacturing.

    Using a RISC-V chip is their newest utopian promise. They might deliver on that, but those in the know among us may have noticed that they will use K1 chips, which are still somewhat buggy and very slow, so its dead on arrival even if they nail every other aspect of the phone.

    It might be possible for a small company to make a good linux phone using the work of past projects, and expanding on them. It might be buggy and unpolished, but it could work.

    It might be possible for a small company to reinvent UX for a phone. It might be buggy and unpolished, but it could work.

    It may be possible to manufacture some form of smartphone in the US. Perhaps with some caveats, but still.

    It may be possible to make a RISC-V phone. It will be slow and buggy, but possible.

    It is completely insane to try all of these things at the same time!



  • If your “anti-abuse” security requires your software to be closed source, then your software is not secure. Well built software can be open source, and still secure.

    You imply that your code could not prevent abuse if it was known, so, no offense, but that code must be pretty bad!

    I like the fediverse because it distributes power. I, as a user, have the power to choose my instance, choose my moderation, choose the structure that my social interactions are built upon (threads based, microblogs, etc) and if I dont like something, I have the power to change it. Thats why open source is such a perfect fit to that model. Closed source federation then, sounds to me like you didnt get it.



  • You know, I was convinced AI software would never be able to build big complex software projects because of its rather sloppy coding style, but times have clearly changed.

    I also worry if the situation turns clone hostile. Just this week, there was an article of somebody reverse engineering their samsung tv using AI(claude I think?), and then they got banned from it! And also this week, Anthropic added a new rule to their harness that disallowed replicating existing games. If this continues, we dont need lawmakers to make this sort of endevour next to impossible. So I suppose: Build it, while they let you!