• 0 Posts
  • 27 Comments
Joined 9 months ago
cake
Cake day: February 5th, 2026

help-circle


  • If you break it into simple tasks, it’s not that bad. For example, for each function, ask it to add reference parameters for the global variables used by that function, and to modify the call site accordingly

    With tasks this small, you’re not asking the LLM to think for you, you’re just using it to automate a boring task

    Once that’s done, seeing each function’s interface should make it easier for your brain to identify what further refactoring is needed






  • We’re talking about Bitcoin (which has no intrinsic value/usefulness) here. I didn’t realize that wasn’t clear, so I have now edited my comment to clarify that

    More gibberish; there are plenty of reasons for crypto currency to continue to gain value.

    Yes, cryptocurrency in general. Not Bitcoin specifically, the most antiquated, slowest, least sustainable one there is (you know, given that it was the first, unlike newer cryptocurrencies which were able to learn from its mistakes and make technological advancements, such as not requiring the power of an entire nation just to run them, and not taking some multiple of 10 minutes before the cashier can finally give you your coffee)



  • glibg10b@lemmy.ziptomemes@lemmy.world•Buy bitcoin, bro
    link
    fedilink
    arrow-up
    26
    arrow-down
    1
    ·
    edit-2
    4 days ago

    Money doesn’t grow on trees. In the case of crypto, it came from the people who lost, or the people who are still in it

    I don’t think there’s any hope for the people who are still in it. This 4-year cycle has already been quite disappointing than the previous ones, and there’s no reason for things to improve from here




  • Why is it an entirely different conversation? Or rather, why do you think that this conversation is not what the AI-opposition are concerned about? (At least some of us are)

    Here are some quotes from the article we all know we’re talking about:

    Torvalds expects contributors to use well-funded chatbot coding

    Kernel maintainers are complaining they’re drowning in slop. Torvalds’ answer is: less people, more chatbots!

    Torvalds said he is not trying to push non-chatbot users out of the project. But there’s already so much slop code and so many slop procedures that his actions don’t match these words.

    These are just from the first few opening paragraphs. I could keep going.

    Now here are some of the top Lemmy comments:

    The most disappointing aspect of using AI on linux, to me, is that it sort of implies linux is now more comfortable with the “move fast and break things” -mindset, now caring more about “efficiency” and “speed” rather than integrity and quality.

    Its really sad, Linus used to be a champion of open source and now hes doing his best to ruin it.

    Linus is a traitor to Open Source community. Using and promoting a tool created by stealing the works of thousands of developers, artists etc. He might be a good programmer but he is a fucking douchebag.

    So yes, the AI opposition absolutely is concerned with the maintainers welcoming slop into the codebase. This is why we needed someone like @auzy1@lemmy.world to step in and make a post clarifying the situation.



  • glibg10b@lemmy.ziptoFediverse@lemmy.world•Hello from a small new instance
    link
    fedilink
    English
    arrow-up
    6
    arrow-down
    1
    ·
    12 days ago

    That was a mistake, and it’s been swapped.

    The Photon-y feel is real: Photon was the client I replaced, so it set the bar for what mine had to do.

    The code is written from scratch; none of Photon’s is in it.

    I’m very familiar with this writing style. Sorry, but I don’t see why I would use a vibe coded instance when there are already so many non-vibe coded instances to choose from






  • In the past, when I saw the same posts on both Reddit and Lemmy, the Lemmy comments always showed more critical thinking and skepticism where it was appropriate. Nothing was taken at face value; memes were not considered a source of information, and there were always many comments looking for verifiable sources. This was in stark contrast to Reddit.

    But, apparently, something changed. When the memes about the bombings started coming in, every comment was based on the premise that the meme was true. “2/3s??? Wow, I was expecting a big number but not a sizeable majority” “reddit too, all the traffic has been quiet.”

    Fortunately, as time went on, the more critical comments got voted to the top, and this spurred more skepticism. But the damage had already been done. The first wave of comments is always more representative of peoples’ uninfluenced reactions; seeing them made me lose my faith in Lemmy.