Lemmy account of natanox@chaos.social

  • 0 Posts
  • 22 Comments
Joined 7 months ago
cake
Cake day: October 7th, 2024

help-circle
  • Like I said it’s less of a problem with KDE, they even got a button to add Flathub specifically in Discover. It’s more of a thing with Gnome and Gnome Software where no “Add Flathub” button exists (and also no GUI to add repos -> they have to look up the whole CLI command), so newer users won’t necessarily be aware that something rather important is missing.


  • Which Nvidia driver setup do you use? The problems arise with the proprietary driver; if you roll back or use a different kernel than the current default (as specified by the repo) both my brother and I had the unfortunate situation of the driver kernel module missing. Nouveau or NVK probably don’t cause such issues.


  • No matter which OpenSuse people end up choosing, it’s a super solid decision. Even though it relies on infrastructure by SUSE S.A., a company that unfortunately has ties to the US (mostly hosting with offices and employees in the US) but got its HQ in Europe, it’s the most solid and user-friendly distro out there if you look for rather independent distros (the only user-friendly one that’s fully independent would be Mageia, but that one really isn’t where it would have to be imho). And the existence of bootable snapshots in case something happened is extremely useful. The biggest problems I’ve found are just 2: Problems with the Nvidia driver (especially if you use said snapshots), and Flathub not coming preconfigured (not a Problem in KDE since there’s a button new users can stumble over, but for Gnome you have to know something rather important is missing to look up the command to add it since there isn’t a GUI to add Flatpak repos yet).

    Other than that the whole OpenSuse ecosystem is just great.


  • Interesting moral question here:

    Given the huge problems are power consumption, morals behind training data and blind trust in AI slop, do you think there is a window of acceptable usage for LLMs as locally run (on existing hardware) coding assistant (not executive tool that does it for you) to help with work on FOSS projects (giving back to where it has taken from) with no money flowing to any company (therefore not bolstering that commercial ecosystem)? While this obviously doesn’t address the energy consumption during training, it may alleviates moral issues to the point people start to think about it as acceptable tool.

    To make it abundantly clear, this is neither about “vibe coding” where it does code for you badly, and definitely not about any other bullshit like generative “art”. It’s about the question of humble, educated use of a potential useful tool in a way it might be morally acceptable.



  • Is this a joke? Even if you ignore the overly fancy stuff like those Forum-like features Matrix can hardly do the basics correctly. The encryption constantly causes issues, Voice feature is still beta, the UX is a mess, the UI is lackluster, basic features are missing (you can’t even set the Voice Activation level, wtf), there are no integrated admin tools and the third-party UIs are either wonky or lackluster, the software (both server and client) is a bloated mess that takes aeons to do anything and is awful to develop (a friend looked at it and quickly decided it’s not worth all the hassle)…

    Of course there’s a difference in size, still they could’ve figured things out way better in the last decade. People ask for voice channels since 2017, and not just did they have a working Jitsi integration, there also are WebRTC frameworks ready to use they could’ve picked. And even now with that Beta feature it is more than obvious they do not want it to work as simple as Discord, but more like a professional software for meetings (that or they just really love creating convoluted UI).

    There’s no way to get any majority of Discord users to use this mess. We discussed this once in our local hackerspace, and while the general opinion of course was more complex given some people know how complicated the protocol is, there also was a consensus that the software as it is is “not great to use”. And if even hackers / enthusiasts are saying this there’s no way in hell to convince casual users.




  • Nothing comes close to the feature set offered by Discord, Matrix’ bad priorities unfortunately made sure of this. There currently is a project to fix their shit even if it means to break some bad decisions, Tuwunel, however it’s neither ready nor is it clear where it will end up. The previous project it forked, Conduwuit, got bullied into giving up.

    You’ll probably be best off with a classic 2-way approach for now. Stick with Mumble for Voice and get something nice for Chat and Organizing like Mattermost or Revolt (or even IRC if you’re a purist). With some luck Discord’s strong enshittification will give projects like Tuwunel the necessary push it needs to force Matrix to finally care for more than just the needs of governments and their perfectionism that gets them nowhere for years now. That or we’ll see some kind of soft-fork with even more bad blood.


  • Yeah… I’m quickly reaching the point where I’m quicker thinking and writing Python code than even writing the prompts. Let alone the additional time going through the generated stuff to adjust and fix things.

    It’s good to get a grip on syntax, terminology and as an overly fancy (but very fast) search bot that can (mostly) apply your question to the very code that’s in front of you, at least in popular languages. But once you got that stuff in your head… I don’t think I’ll bother too much in the future. There surely are tons of useful things you can do with multimodal LLMs, coding on its own properly just isn’t one of it. At least not with the current generation.



  • Yeah, same with Codestral. You have to tell it what to do very specifically, and once it gets stuck somewhere you have to move to a new session to get rid of the history junk.

    Both it and ChatGPT also repeatedly told me to save binary data I wanted to store in memory as a list, with every 1024 bytes being a new entry… in form of a string (supposedly). And the worst thing is that, given the way it extracted that data later on, this unholy implementation from hell would’ve probably even worked up to a certain point.




  • Natanox@discuss.tchncs.detoMemes@lemmy.mlRednote right now
    link
    fedilink
    English
    arrow-up
    2
    arrow-down
    3
    ·
    3 months ago

    Now if we could just get rid of all those power-hungry dicks who keep poisoning the human spirit, let it be with Capitalism, Maoism, Neoliberalism and whatnot…

    Thing is, the concern about RedNote is completely valid. Even if you strip away any overarching US propaganda or whatever, we know that the CCP does really, really shitty things as well and take heavy, manipulative influence (yes yes, the US and its defacto Feudal Lords owning Meta etc. do so too).

    I’m just angry people won’t take their time to look into this whole topic and go with more propaganda-resistant, federated alternatives. Not perfect, but better than a centralized service that can be influenced by either China or the US. Or Russia… or literally anyone. So that we can, in the end, just dance to some nice human music.






  • Like every time a country conducts genocide people close their eyes and don’t want to admit that, facing such crimes against humanity, every single Israeli has the duty to do whatever they can to stop their government. No matter if its a Hitler, Stalin, Mao, Jinping, Putin or Netanyahu (or some political conglomerate), and no matter where the injustice and moral corruption began or what once was, the people have to stand up against a corrupt system at that point. Of they don’t they are indeed complicit, every single time.