• 0 Posts
  • 30 Comments
Joined 2 years ago
cake
Cake day: June 6th, 2023

help-circle

  • I own a OnePlus 6 with postmarketOS. My daily driver is a Pixel 7 with CalyxOS and microg turned off.

    Despite having effectively only FOSS apps on my Android daily driver, I can’t daily drive postmarketOS. It’s making great progress, but isn’t nearly stable enough as a modern smartphone, and several other issues hold it back;

    • Sleep states. Currently you get to choose between your phone going to sleep (~ a day battery life) but without notifications, incoming calls, alarms, etc. Or your phone “staying awake”, where you’ll have those features but only ~4 hours battery.
    • Hardware video decoding, which “can work” but only in select apps, and they’re not great for mobile.
    • Audio issues, such as no audio from the earpiece, microphone not working, or no audio at all.

    If you rely on non-foss Android apps, there is Waydroid, but it’s not a perfect solution and might have issues.

    It’s not a “waste of money” if you want a device to experiment or tinker with, or if you want to follow progress of Linux mobile, but it is extremely unlikely to replace your daily driver.


  • deadcade@lemmy.deadca.detoMemes@lemmy.mlFirst law of chat robotics
    link
    fedilink
    arrow-up
    4
    arrow-down
    1
    ·
    2 months ago

    It’s still an LLM, not a “truth machine”. Replying with “did you make that up” will just cause it to respond with the next most likely tokens.

    Try this: if you know it’s saying something factual, try your technique. It will likely “correct” itself by slightly rephrasing. Enough rephrasing might change the meaning of the sentence, but there’s nothing checking whether that’s factual before or after.

    I’ve had some LLMs become extremely stubborn, and deny that it’s wrong on basic facts like the release year of certain media.


  • None of that’s true. Free speech laws try to prevent the government from arresting you for opinions or criticism. Social media platforms, parents, etc are still able to take action against statements without reason. The government can also put the blame on something else. If someone is critical of the government, they’re likely to have broken laws they don’t agree with.







  • Client side anti-cheat is inherently flawed. These games are asking an untrusted computer whether it is cheating. That’s like asking a known liar whether they’re lying at that moment. The one way to make it harder for the computer to “lie” is by increasing the permissions the AC has, which comes at the cost of privacy for people with the game, and security for every Windows user (not just the ones with a certain game installed).

    Client side anti-cheat can be poked and investigated locally, with no restrictions. All it takes a skilled enough cheater is time, and they will bypass it. The only way to test server side anti-cheat is by hopping in the game, trying to learn how it works, and trying to bypass it. That is a much more time consuming and expensive process.




  • Surprising that Boneworks wasn’t mentioned. The whole game is physics based puzzles, meaning you can either solve them, or stack a couple boxes and jump really high. These types of solutions are encouraged in the game, and there’s a couple puzzles I’ve never even solved because the walls were too low.




  • On the Steam Deck, while using SteamOS (or other Linux distros), EAC (and a few others like Battleye) run in userspace, not as kernel level.

    The intention of Anti-Cheat and DRM is to hide what they’re doing, in an attempt to prevent people from cheating or pirating. Malware often uses similar techniques to hide what it’s doing.

    Kernel level Anti-Cheat runs with the highest level of permission on your system, meaning it has access to everything happening on your PC, and all your hardware.

    That means kernel level Anti-Cheat can do whatever it wants on your computer, and it’s intentionally hard to figure out what it’s doing. Even though it’s probably not harmful, it shares a lot of similarities with actual malware, and we can’t be fully sure whether it is harmful or not. This is why a lot of people are against kernel level Anti-Cheat.

    EAC, afaik, has acted as just an anti cheat, and is therefore likely not harmful to your system. However, like other Anti-Cheats, it is harmful with the standards being set.


  • That’s, very odd. Just watched the bit about TP2 in the video, and I’m getting nowhere near that on my Steam Deck (non-oled).

    Setting everything to low, and FSR performance, it looks awful. There’s very obvious upscaling artifacting, especially during motion. Performance is playable at about 30-40 fps, except in the “starting hub” of the game, where performance can dip to 10fps at times (although no real gameplay occurs there).

    With everything being set so low, LODs are clearly visible, even on the small screen. Gaps in signs, thin walls, and stairs are visible from ~5 in game meters away.

    On the settings they show in the video, with a fresh save, I get similar numbers in the first couple minutes of the game, but FPS tanks after that. On a save further into the game, I’m getting maybe 20fps (50 when staring at the floor).

    The game is still very playable on the lowest settings, and if you’re into puzzle games like The Talos Principle, it’s still a good experience. I’m not normally one to stream my games from my PC, but The Talos Principle 2 is just a better experience with more powerful hardware.



  • Most malware is written for Windows, especially when it’s distributed as a Windows executable. (Almost) no Windows malware targets Wine specifically. However, Wine on its own is not a sandboxing tool, and Windows ransomware will ruin your day.

    Bottles does two things for security:

    1. Separate wineprefixes
    2. Being a flatpak

    By separating wineprefixes, as long as the host filesystem is not directly exposed (which iirc is default for bottles), any malware not written with Wine in mind will only affect its own “bottle”.

    By being a flatpak, even if some Windows malware specifically targets Wine, it would still have to escape the flatpak sandbox for elevated permissions. If the bottles flatpak has no access to personal files, “Wine-aware” malware won’t either.

    Although malware can still do damage, even in its own sandbox. For example, botnet type malware would still function. The host system is “safe”, but the damage can still be done externally. Usually application-defined “autostarting” of applications is broken under Wine (iirc), which means all non “Wine-aware” malware will only start when an infected windows application is started in bottles.

    Any sandbox will eventually be escaped, and malware sophisticated enough will be able to get access to everything on the host system. The chances of running into malware like this in the wild are extremely small.

    • Is it fully secure? No.
    • Is your virtual Windows environment safe? No.
    • Are other “bottles” safe? Likely, as long as the malware isn’t aware of Wine.
    • Is your Linux host safe? Most likely, depending on your flatpak settings. (and the malware has to specifically target Wine under Flatpak).