• 0 Posts
  • 20 Comments
Joined 10 days ago
cake
Cake day: April 20th, 2026

help-circle


  • Yeah, the big thing is that management has no sense how little coding you actually do in a software engineering role. You spend so much more time understanding requirements, understanding how you can resolve roadblocks within your organization and understanding what the hell the code does that was previously written.

    In particular, the last part is something that will most definitely take longer for vibecoded programs.
    The code is often needlessly complex, because:

    • folks throw in additional features with no restraint,
    • the AI will gladly generate a second implementation for stuff, you already solved in the codebase, and
    • AI-generated code tends to just be noisy, because you need rigorous logical reasoning to find the most minimal solution.

    But you also just don’t have human beings that made all the detail decisions and can tell you why they’re important. In vibecoded code, all of these detail decisions are accidental and only ‘proven’ in so far as the given accidental state that the code is in, happens to not explode in reality. If you need to tweak anything about it, you’re completely blind as to what’s actually important and what’s just in there, because the AI figured, it’s the most likely thing to autocomplete there.


  • I mean, even then, they could increase the price per token, if they want to hand out fewer tokens for the price paid.

    They could make this work like a prepaid SIM card, where you charge it with e.g. $10 and then you can use it until the $10 are used up.
    Instead, they make it work like in-game currencies in scammy free-to-play games. Except that they didn’t choose a confusing conversion rate, for some reason…



  • Yeah, I imagine that they did try. But it’s not just the intentionally misleading announcement post, they also have 5(?) different subscription tiers, which get different changes from this. And one of the subscription tiers is actually called “Pro+”, so that does not mean “Pro and more expensive tiers” like I wondered. And they have this ridiculous intermediate currency to make things even more confusing.

    Their offering itself is overly complex and confusing…





  • So, did they use AI tools to type “LGTM” 400 times or nah?

    But yeah, I also find that frustrating. Management just looks at terrible metrics like PRs closed or lines of code produced.
    It’s not even novel that you can produce terrible code very quickly. Decades ago, our industry learned that it isn’t worth it, because you suffer for it later. Now the game is altered slightly and management demands that we throw all these learnings out the window.



  • Personally, I found it worth playing around with. I cared less than I thought where I had to move my eyeballs to, once I didn’t have to make the decision anymore.

    And automatic tiling can also enable workflows that just don’t make sense with manual tiling, for example master-stack-layout where basically one window takes up half the screen and the other windows share the other half, and then you swap out which one’s the big window as you see fit.

    But I also wouldn’t have written all that, if I didn’t have a way that you can easily try it out: You can add automatic tiling into KDE Plasma via Kwinscripts. Personally, I’m using Krohnkite: https://store.kde.org/p/2144146
    (Easiest to install by going through the System Settings…)





  • I find a gameplay goal more important than a story goal. Sandbox games like Luanti are tricky for me, because I need to decide what to do with no real reason to do anything. But if a roguelike tells me “There’s an artifact at the bottom of this dungeon. Good luck!”, that’s already more story motivation than I need, because the gameplay goal is straightforward.

    I also find lots of story motivations terrible to begin with, though, when it’s basically “You’re the hero! Go save the world!” and then the gameplay is just genocide. I don’t care, if we’re violencing pixels, but specifically the attempt to justify this violence, is almost always distasteful.





  • For what it’s worth, when we say we do TDD in my team, we write a singular test case that fails, then we implement the production code until the test case works. Then maybe do a bit of refactoring to make it all work nicely together, and only then you start with the next test case.

    Writing swathes of unit tests upfront sounds absolutely mad to me, for the reason you state. But also because you do need an API to test against. You can’t write a unit test in complete isolation, pretty much by definition. You can often do so for integration tests, but you definitely don’t want to put all test cases into integration tests, as that increases complexity massively…