Skip Navigation

spartanatreyu

@ spartanatreyu @programming.dev

Posts
14
Comments
308
Joined
3 yr. ago

  • Just saying

    Jump

  • Ironic

    1. The video:

    • The video literally goes over the numbers in worst case scenario for solar and still comes out ahead, while also going through a bunch of mistakes that people make and deconstructing the gotchas along the way.

    1. Your comment:

    • Saying you weren't going to watch the full video
    • Skimmed some of it and took some numbers and made mistakes

    1. Someone else:

    • Links to the video you skipped that would have gone through your mistakes before you commented.

    1. Your response:

    • This is the same video

    YES! It already goes over the mistakes you were making.

    On a side note, this was probably the best video I've seen in the last 12 months.

    I thought it would be a nice and nerdy breakdown of solar panels, but the more I watched the better it got.

    For those who did watch it: wow what a whiplash!

  • Node.js vs Deno vs Bun Performance Benchmarks

    Jump

  • Bro, how are generators going to be faster?

    This is an AI article.

    My results:

    Firefox:

    Loop: 44ms - timer ended

    Generator: 4580ms - timer ended

    Node (uses same engine as deno, chrome, edge):

    Loop: 30.577ms

    Generator: 1.533s

    Safari (uses same engine as bun):

    Loop: 605.222ms

    Generator: 2804.669ms

    Bun (same engine as Safari but without needing to apologise for Safari):

    [17.52ms] Loop

    [297.17ms] Generator

    Generators are going to be slow because:

    • they're going to be stack switching so much in JS runtimes which adds a lot of overhead
    • JS doesn't have the other language features (yet) that you want to use with generators, which makes less folks want to use generators, which makes implementers not want to spend time optimising them. (Why bother trying to inline generator state when it's probably going to change once the adjacent features come in?)

    Until generators don't rely on stack switching, they're always going to be super slow.

  • Node.js vs Deno vs Bun Performance Benchmarks

    Jump

  • This benchmark is pretty useless.

    There's no benchmark code shown to see if they're doing something wrong or cherrypicking.

    Also, they only tested on arm64? Why weren't they testing on x86-64?

    And what the hell is this test?

    Common data transformation pattern, map then aggregate with reduce.

    People who care about performance are using loops, not map(). Why are you even testing the slow path?

  • Man jailed for a year after endorsing neo-Nazi views and making antisemitic speech at Sydney rally on Australia Day

    Jump

  • Tolerance is only paradoxical when you go out of your way to not view it as a social contact.

    When tolerance is the social contact, then everyone is protected by it except those who go out of their way to not be protected by it.

  • CIOs told: Prove your AI pays off – or pay the price

    Jump

  • The sound is so long and drawn out, it's more of a whoopie cushion sound

  • How do you pronounce daemon?

    Jump

  • I say daymon not to avoid offence, but since it sounds cooler than demon.

    A demon sounds like a fiend that has only been around for at most a few hundred years, but a daemon sounds like it has been around for a few thousand so it is much more dangerous.

  • How have you made art with code?

    Jump

  • I'm usually programming, but I've got a few bits and pieces that can be shared:

    The rest are too experimental, or are released commercial projects.

  • "your website doesn't work without javascript"

    Jump

  • Let me load HTTP without the S if I want to.

    No, that let's companies man-in-the-middle you.

    ISPs literally couldn't help themselves inject ads and other scripts that lagged and broke everything on every website all to chase a few bucks.

    HTTPS prevents them from doing that.

  • Experimental Zones Protocol Merged To Wayland After 2+ Years, 620+ Comments

    Jump

  • It's Phoronix, the articles are good but the comments on every article are unbearable.

    It's basically a meme at this point.

    There's a bunch of people in the forum with poor awareness and emotional control throwing tantrums, typically over any project with a code of conduct (even something as simple as "Don't be an asshole") because they believe it infringes on their right to be inconsiderate.

    New projects want to avoid the drama and issues of projects from decades past, so naturally they add codes of conduct which causes those folks above to have hissyfits everywhere about those projects anytime they or something adjacent to them is mentioned.

    Most places warn then kick those users out, and then there's Phoronix.

  • COSMIC Epoch 2 and 3 Roadmap

    Jump

  • This is the kind of divorced from reality slop I expect to see from managers who think that 9 mothers can birth a child in one month

  • COSMIC Epoch 2 and 3 Roadmap

    Jump

  • I don't understand comments like this

    It’s a wishlist of Open tickets. Wouldn’t necessarily even call this a commitment to a roadmap.

    Roadmaps are just wishlists with a committed to priority list, like an assembly line. Programming doesn't work like an assembly line and anyone trying to convince you to the contrary is bullshiting marketing.

    Every task has the potential to bring forwards new information that changes the order in which tasks should be completed, let alone which tasks even need to attempted. You only get all the information once you've finished.

    Instead of having an inflexible commitment, use a wishlist, and do the tasks in order of which one makes the most sense at the time.

    75% of Open tickets will never get resolved anyway.

    Based on what?

    You can clearly see that practically all tickets in their previous epoch were resolved: https://github.com/orgs/pop-os/projects/23/insights?period=max

  • Rust Coreutils Continues Working Toward 100% GNU Compatibility, Proving Trolls Wrong

    Jump

  • If I were a uutils developer, I would stay far away from all of these discussions because of how much hate is directed towards it.

    I only see the hate towards this project being either from anti-rust trolls, or misdirected hate from Ubuntu towards switching to a new coreutils implementation on an LTS release before full compatibility has been achieved. I don't see any hate in regards to licensing.

    If they do not adopt the license you prefer, would it be better for them to just go ahead and abandon the whole effort? Are there efforts really so valueless simply because they chose the license that they did?

    Their efforts have value, but the value is limited by its current license. MIT licenced projects have a recurring history of being improved privately without those improvements going back into the project. It leads to a lot of duplicated, wasted effort. There may also be the potential for patent issues with the licence. No one wants to deal with some litigious asshole or company going after the project turning it radioactive.

    Moreover, is dictating to volunteers what license they should be using for their code what you think this community should be about?

    I think bringing up issues with the project is definitely something that should be brought up. As for dictating which particular licence is used, that's up to the contributors, but that doesn't mean others can't give their input. It's also likely that most of the contributors will want a license that allows the project to safely continue into the future.

    You claim that it is important that people make tons noise in every single post on uutils because it will prevent a bad scenario down the line, but could you detail what that scenario is? Because people like to make allusions to such a scenario constantly but refuse to get specific and then engage on a discussion on the specifics.

    I thought the Redis example was a good example of this.

    I continue on this point further down, but I'm leaving this right now to stay on topic with redis.

    Incidentally, your choice of Redis is an example exactly illustrates my point that this license is not a gigantic deal it shows that the worst case scenario is… uutils being forked.

    The community was fractured. A report by an enterprise support company said 75% of existing redis users were motivated to seek alternatives. I'm not sure what number you would consider to be a gigantic deal, but Redis certainly thought it was, otherwise they would not have reverted back to the previous license.

    Heck, it can even be forked at any time with a copyleft license precisely because its existing language is permissive.

    It can be forked, but relicensing can mean needing permission from every contributor of the original, and/or removing all contributions from those who don't agree to the new licence. Not to mention the community fracturing, and legal issues. It's a massive effort that can be prevented by the original project choosing a better license earlier.

    You claim that it is important that people make tons noise in every single post on uutils because it will prevent a bad scenario down the line, but could you detail what that scenario is? Because people like to make allusions to such a scenario constantly but refuse to get specific and then engage on a discussion on the specifics.

    Well this comment is probably getting too long, so I'll simply point you towards the busybox licensing drama.

  • Rust Coreutils Continues Working Toward 100% GNU Compatibility, Proving Trolls Wrong

    Jump

  • So it needs to be commented on in every single article?

    Yes

    If so, is that going to change anything?

    Potentially.

    The alternative is not bringing up the concern and it goes forgotten until it is too late and we are stuck with the results of bad decisions for no good reason.

    Developers voicing their concerns is the only way things can change for the better.

    Here's two examples:

    1. Redis licensing rug pull

      • Redis unexpectly changed its own licencing
      • Developers demanded Redis return to its original (or similar) licence
      • Redis said no
      • Developers formed their own Redis clones from scratch with compatible APIs
      • Developers switched to the new Redis replacements
      • Redis returned back to the original licence in an attempt to keep existing users

    2. Google's JpegXL whiplash

      • Google added support for jpegxl in Chrome
      • Google removed support for jpegxl in Chrome in favour of inferior standards
      • Developers demanded support added back
      • Google said no
      • Developers flooded every issue tracker and feature request with support for jpegxl, consistently, unrelentingly, for years
      • Other browsers add support for jpegxl
      • Creative industry adds support for jpegxl
      • PDF association adds support for jpegxl
      • Google forced choose between jpegxl or fall out of supporting pdf standard
      • Google readds jpegxl support

    And there's plenty of other examples (e.g. Microsoft against linux -> WSL support, etc...)

    If developers don't voice their concerns, then things stop changing for the better.

  • Rust Coreutils Continues Working Toward 100% GNU Compatibility, Proving Trolls Wrong

    Jump

  • Licencing is a legitimate concern (not noise), even more so considering it's for the core utils

  • Locked

    Yes, Windows is the Problem.

    Jump

  • macOS has far better desktop applications across the board than Linux. What you get for free with a Mac is nothing to sneeze at:

    Counterpoint: The apple of yore is not the apple of today.

    Recent examples:

    At this point, the beginner-friendly linux distros (Linux Mint, Ubuntu, Kubuntu, Bazzite) are more sensible and logically consistent than either Windows or MacOS.

    Tangent: If you're looking at paid git clients on MacOS, I'd recommend fork over Git Tower or Kaleidoscope.

  • KORE Programming Language

    Jump

  • Significant whitespace, minimal ceremony

    Well that's an oxymoron. If you're having to alter the whitespace instead of the code because the whitespace is significant than all you have is ceremony.

  • Help open the source of the myGov Code Generator app

    Jump

  • Not if the standards rely on dedicated hardware, for example: security enclaves.

    The whole point of using a security enclave on different hardware is so that its memory cannot be accessed or polluted by a process or side channel attack on the main hardware.

  • Help open the source of the myGov Code Generator app

    Jump

  • Counterpoint: Public funds pay for software used by military and intelligence services. Certain information becoming publicly available can lead to real harm. (e.g. A self-assessment on a country's own weaknesses, methods that spies deployed abroad can deliver information back, etc...) How do you manage the infohazard risk?

    all software paid for by public funds should be open source.

    Should probably be changed to: all software paid for by public funds should be open source so long as their is no or low foreseeable infohazard risk.

  • Help open the source of the myGov Code Generator app

    Jump

  • Counterpoint: A government portal needs to be extremely backwards compatible to support as many people as possible. That includes supporting devices that don't support the latest standards.

  • Wayland didn’t kill Linux desktops, but it did expose their weakest assumptions [EDIT: Warning, article likely written with AI assistance]

    Jump

  • That should be:

    People use used windows because it has had decades of backwards compatibility sort of.

    Nowadays: it has current compatibility sort of.

    Microsoft is a mess now, so it's no surprise that their current Windows version is also a mess.