Microsoft has long wanted to get vendors out of the kernel. It’s a huge privacy/security/stability risk, and causes major issues like the Crowdstrike outage.

Most of those issues also apply to kernel anti-cheat as well, and it’s likely that Microsoft will also attempt to move anti-cheat vendors out of kernel space. The biggest gaming issues with steamOS/Linux are kernel anti-cheat not working, so this could be huge for having full compatibility of multiplayer games on Linux.

  • kadu@lemmy.world
    link
    fedilink
    arrow-up
    1
    ·
    edit-2
    13 days ago

    This is what, the fourth time a Linux community gets excited about this? But that’s actually not good for us at all. Much like Android’s safety net, or the nightmare that is the Mac equivalent, the entire point will be creating an untouchable chain from the firmware to the final OS being booted, and only allowing some apps to use a specific API to attest this isn’t compromised.

    This is horrendous for people trying to modify the OS or, in a more relevant tone, run programs meant for that OS on an entirely different environment. Microsoft has slowly been moving towards making this work on PCs, mostly due to pressure from DRM providers like Netflix or banking apps, but unlike Apple they can’t simply lock everything down at once and say “deal with it” because Windows lives by backwards compatibility. Either way, this is just another step towards this upcoming future.

    If your favorite games now start asking Windows if the chain of trust is not tampered with… say goodbye to compatibility with Proton.

    • MoogleMaestro@lemmy.zip
      link
      fedilink
      English
      arrow-up
      0
      ·
      edit-2
      13 days ago

      Yes,

      but game companies also want to spy on you and potentially sell your data. Even if they aren’t selling it, the ability to do so increases the value to investors. This is the way tech companies talk about invasive software in general, FWIW.

        • adr1an@programming.dev
          link
          fedilink
          arrow-up
          1
          arrow-down
          1
          ·
          10 days ago

          Data brokers do exist… Who they buy from is the only privacy they respect. You know, capitalism.

          • Derin@lemmy.beru.co
            link
            fedilink
            arrow-up
            1
            arrow-down
            1
            ·
            10 days ago

            So by that logic we shouldn’t be downloading any precompiled binaries from the net - they could all be spying on us!

            • adr1an@programming.dev
              link
              fedilink
              arrow-up
              1
              arrow-down
              1
              ·
              edit-2
              10 days ago

              a binary and a kernel module are not the same. And I was talking about business practices that are known. But nevermind, that was before I understood you are just trolling. Now I’ll simply wish you a wonderful reddit experience.

              • Derin@lemmy.beru.co
                link
                fedilink
                arrow-up
                1
                arrow-down
                1
                ·
                10 days ago

                Not trolling at all. I’m a game developer, so I was curious to hear about instances of game devs using kernel level anticheat to harvest people’s personal (and identifiable) data to sell to data brokers.

                Glad to know there aren’t any examples of it outside of people screaming about capitalism - which is, let’s be honest, quite indicative of the Lemmy experience these days.

                • adr1an@programming.dev
                  link
                  fedilink
                  arrow-up
                  1
                  ·
                  edit-2
                  10 days ago

                  As for actual kernel anticheat software siphoning out data, again, it’s something that we should not immediatly dismiss… The lack of evidence is not entirely on either side of this sudden ‘you vs. me’

                  Acxiom does sell a package of “Gaming” data. Probably coming from mobile phones for the most part, since that’s where most Studios are more aggressive (even towards children, see Tilting Point Media LLC settlement last year)

                  Again, knowing which Studios are selling data (identifiable or not) is impossible if no Court interferes.

                  I was only trying to make the point that it is feasible… That’s why I referred you to recall that “capitalism bad yadda yadda…”

  • kieron115@startrek.website
    link
    fedilink
    English
    arrow-up
    0
    ·
    13 days ago

    I’m curious to see how CompTIA responds to this. They already don’t allow you to take their exams in a VM or any kind of Linux. Presumably for the same “concerns” that the anti-cheat industry has.

      • kieron115@startrek.website
        link
        fedilink
        English
        arrow-up
        1
        ·
        edit-2
        11 days ago

        As a holder of multiple CompTIA certificates I wholeheartedly agree that they’re useless. Unfortunately they’re by far the most common means of contractors (the actual people, not the companies) checking off the boxes to qualify for U.S. government IT contracts; which means they’re still relevant.

  • 🇰 🌀 🇱 🇦 🇳 🇦 🇰 🇮 @pawb.social
    link
    fedilink
    English
    arrow-up
    0
    ·
    edit-2
    13 days ago

    I’d probably be okay with kernel level anti-cheats if they actually stopped cheaters. But they don’t. Hell, the best anti-cheat I’ve ever seen that actually works isn’t even made by the developers of the game; it’s a mod! Blue Sentinel for Dark Souls 3. All it does is check if the files a player you’re connecting to has deviate at all from your own, then prevents the connection if they are not 1:1 identical.

    • warm@kbin.earth
      link
      fedilink
      arrow-up
      0
      ·
      13 days ago

      Basic anti-cheat already does this, but also with memory, because most cheats are reading/modifying what is in memory. I think the only ethical solution for anti-cheat is on the server side, with machine learning perhaps, kind of like VACnet.

      • filcuk@lemmy.zip
        link
        fedilink
        arrow-up
        0
        ·
        13 days ago

        The problem is that, with a good enough cheat, it can be impossible to distinguish from a very good player.
        The best cheats use a secondary device emulating human input and reactions, which is practically undetectable.

        • viking@infosec.pub
          link
          fedilink
          arrow-up
          2
          ·
          13 days ago

          A secondary device can’t be identified by kernel level anti-cheat either. If you have a standalone device that identifies as a USB keyboard and mouse and then generates inputs that give you a 100% headshot count, there’s nothing you could detect through the kernel, since all it detects are keystrokes and clicks.