• angel@sopuli.xyz
    link
    fedilink
    arrow-up
    141
    ·
    4 months ago

    Rustdesk looks good on the outside, but if you look inside, it has a really bad codebase and has done some sketchy stuff in the past.

    Last year, it installed custom root certificates as trusted on windows, which is a huge security risk: https://github.com/rustdesk/rustdesk/discussions/6444

    On linux systems, it forced its own autostart with no option to disable this behavior: https://github.com/rustdesk/rustdesk/issues/4863

    In the past, when it didn’t have Wayland support yet, it edited your GDM config and just disabled wayland: https://github.com/rustdesk/rustdesk/blob/1.1.9/src/platform/linux.rs#L411-L422

    Furthermore, the code quality is really bad. 90% of the linux platform-dependant code is just executing shell commands and parsing their output, while the same could be achieved in a safe way with proper rust builtins: https://github.com/rustdesk/rustdesk/blob/master/src/platform/linux.rs

    While I agree that Rustdesk works pretty flawlessly, the codebase and the behavior of the developers made me distrust the software and I don’t recommend using it.

    • boredsquirrel@slrpnk.net
      link
      fedilink
      arrow-up
      42
      ·
      edit-2
      4 months ago

      @petsoi@discuss.tchncs.de you might want to add that warning to the post.

      They also tried to submit the app to Flathub, but had way too broad permissions with no explanation why. “Users expect filesystem access” etc. In the end it was rejected and they publish a .flatpak file themselves.

      https://github.com/flathub/flathub/pull/5233

      The other points are far worse though.

      • Karna@lemmy.ml
        link
        fedilink
        arrow-up
        9
        ·
        4 months ago

        Rustdesk controversy

        The whole discussion on that pull request is extremely sketchy, IMO.

    • Possibly linux@lemmy.zip
      link
      fedilink
      English
      arrow-up
      38
      ·
      edit-2
      4 months ago

      To add on:

      • There is no transparency about who is behind it. It just a Github account called “Rustdesk.” It could be a real company in Singapore or it could be some guy in China as people have speculated.

      • The Rustdesk software needs way more permissions than necessary. This became evident with the flatpak as they did sandbox escapes which prevented them from being on flathub

      • The Rustdesk distribution is entirely centralize release server run by Rustdesk. They could easily push out malware to lots of devices.

      • They have done some sketchy things in the past. One of the things they did was quietly switch Linux desktops back to X11.

      • The Rustdesk system is not terribly resistant to brute forcing. The weak password means they someone could try every combination.

      • Rustdesk docker deployment docker compose exposes all ports on the host. This is minor but it could lead to a sandbox excape.

      • Rustdesk servers keep getting hosted in countries that have freedom problems such as China and Russia.

      • bluetoque@lemmy.ca
        link
        fedilink
        English
        arrow-up
        16
        ·
        4 months ago

        Wow, I’m wondering how anyone would trust this software. It literally exposes your desktop. To me that requires top-tier trust level, i.e. nothing sketchy at all.

      • Fonzie!@ttrpg.network
        link
        fedilink
        arrow-up
        7
        ·
        edit-2
        4 months ago

        could be some guy in China

        I don’t see how that’s a problem, it’s not like it’s by a Chinese run company or like the Chinese government is spying on you; in the case you described it’d just be a rando with a hobby/vision.

        The fact that it keeps getting hosted in countries that have freedom problems, such as China and Russia, does concern me, though.

        • Possibly linux@lemmy.zip
          link
          fedilink
          English
          arrow-up
          2
          ·
          4 months ago

          The problem is that China makes developing privacy and freedom friendly tech illegal. You won’t find many Tor devs in China

          • Fonzie!@ttrpg.network
            link
            fedilink
            arrow-up
            2
            ·
            edit-2
            4 months ago

            How do they make that illegal?

            I can’t find much on tech impeding laws online, whatever search terms I enter related to China and privacy just leads me to articles about their data protection law.

    • haui@lemmy.giftedmc.com
      link
      fedilink
      arrow-up
      15
      ·
      4 months ago

      Okayyyy… thats not great. I just read one of the threads and thats scary.

      The person(s?) maintaining this seems to be VERY BAD at communicating. They did fix the auto start problem but did not at all discuss this from what I see. Thats not great.

    • Lettuce eat lettuce@lemmy.ml
      link
      fedilink
      arrow-up
      12
      arrow-down
      2
      ·
      4 months ago

      Really sad about this, because Rust Desk has been the absolute best remote access tool I’ve ever used in the IT world, and that includes many different professional tools like Ninja& Teamviewer.

      It’s so clean, easy to install and run, fast and low latency, handles multi-monitors great, runs on mobile, Linux, Windows, etc.

      Such a shame that it is mired in controversy.

    • Psyhackological@lemmy.ml
      link
      fedilink
      arrow-up
      3
      ·
      edit-2
      4 months ago

      Wth is that, that is the most anti-idiomatic code I have ever seen

      https://github.com/rustdesk/rustdesk/blob/master/src%2Fplatform%2Flinux.rs#L176

      pub fn get_cursor() -> ResultType<Option<u64>> {
          let mut res = None;
          DISPLAY.with(|conn| {
              if let Ok(d) = conn.try_borrow_mut() {
                  if !d.is_null() {
                      unsafe {
                          let img = XFixesGetCursorImage(*d);
                          if !img.is_null() {
                              res = Some((*img).cursor_serial as u64);
                              XFree(img as _);
                          }
                      }
                  }
              }
          });
          Ok(res)
      }
      

      I’m not an expert but this seems wrong.