I’m the Never Ending Pie Throwing Robot, aka NEPTR.

Linux enthusiast, programmer, and privacy advocate. I’m nearly done with an IT Security degree.

TL;DR I am a nerd.

  • 0 Posts
  • 27 Comments
Joined 25 days ago
cake
Cake day: November 20th, 2024

help-circle



  • N.E.P.T.R@lemmy.blahaj.zonetoSelfhosted@lemmy.worldMy thoughts on docker
    link
    fedilink
    English
    arrow-up
    3
    arrow-down
    1
    ·
    edit-2
    3 days ago

    Docker is good when combined with gVisor runtime for better isolation.

    What is gVisor?

    gVisor is an application kernel, written in memory safe Golang, that emulates most system calls and massively reduces the attack surface of the kernel. This is important since the host and guest share the same kernel, and Docker runs rootful. Root inside a Docker container is the same as root on the host, as long as a sandbox escape is used. This could arise if a container image requires unsafe permissions like Docker socket access. gVisor protects against privilege escalation by only using root at the start and never handing root over to the guest.

    Sydbox OCI runtime is also cool and faster than gVisor (both are quick)



  • In that case it is a ToS violation, not piracy. You aren’t paying anything, nor does google lose any money since they have been already paid. We would have to stretch the definition of piracy to include other ToS violations since it is not a financial lose.

    Let’s extend the scenario. If YouTube ToS required you to click every ad to use their service, would it be piracy if someone doesnt follow those instructions? I think it would be a ToS violation, but what damages could Google even seek?

    I hear people sometimes mention that “Google needs to pay somehow to keep YouTube running.” I have no sympathy for Google since they conspired to intentionally push out other video hosting platforms to create monopoly on the market. It is their own fault that videos aren’t more spread out among providers.

    How would you even pirate YouTube anyways?


  • I recommend Mull. It is security/privacy hardened Firefox and built by using Fennec as a base. Always use Fennec over Firefox because it removes telemetry, proprietary code, and strongly protects against browser fingerprinting. Comes with support for most (if not all) desktop Firefox extensions. I highly recommend using uBlock Origin, ask anyone and they’ll tell you it is the best content blocker available.

    Another good browser is Cromite. It is security hardened Chromium with built-in ad/content blocking, decent fingerprinting protection, and strong site isolation. It doesn’t have support for extensions because upstream Chromium on Android doesn’t either and it is hugely complex to port.





  • The weakest link of any secured system is the user. I know that will never change, especially as computers/software become more complicated over time. But I don’t understand why many people argue that “since the user is the weakest link, we don’t need more secure systems, we need better users.” We need both.

    Explaination

    For anyone who suggests that a user can “just be smarter and not install malware” think about this: do you check read all the commits to the software you install, for each update, and then compile from source. The answers is no. And I don’t think we should need to.

    Linux is not secure, it is still meant for tinkerers and by design is very open. This is one of my favorite aspects of Linux, just how open it is. The result though is an insecure system with many attack vectors that are hard to protect against.

    For example, I recently wanted to patch a game for mod support. This required me to run a script that i didnt fully understand. I did my best to read it and nothing looked suspicious, but I couldn’t fully understand because I am not a modder for that game.

    This script could have done a number of things:

    • Added a fake sudo script to the path in the user’s home resulting in privilege escalation.
    • Created a user Systemd service that logs everything added to the clipboard or keylog, since that is also possible on Wayland with an LD_PRELOAD attack.
    • Create a Systemd service that records the screen to grab passwords.
    • Edit the user’s .bashrc file.
    • Delete/encrypt every file owned by the user.
    • Read and exfiltrate all app data from the $HOME
    • Or a combination of multiple other things.

    The solution is sandboxing, permission system, secure defaults, and transparency to the user. And of course a way to disable security checks for tinkerers.

    My point is that the perfect user does not exist. We (inevitably) use our computers to do all sorts of niche things, the perfect user does not even turn their PC on.




  • N.E.P.T.R@lemmy.blahaj.zonetoLinux@lemmy.mlDo you use Gnome or KDE Plasma?
    link
    fedilink
    English
    arrow-up
    6
    arrow-down
    3
    ·
    8 days ago

    I would agree but it still heavily relies/requires X11 for many of the core components and apps. Wayland support on the horizon. For now though I wouldn’t recommend anyone use any DE that doesnt support Wayland, since X.Org is has been unmaintained for many years, it is a bloated protocol, and is insecure by design. This is because it was designed 40+ years ago in a time before security was a big concern to developers.



  • Canonical, the owners of Ubuntu, love to steal open source projects. They’ll help a project with development power, then force the contributors to sign a CLA (for an example see the fork of LXD called Incus). Canonical also uses and forces proprietary systems onto the user’s, e.g. Snap uses the proprietary and hardcoded Canonical repository, which Ubuntu now defaults to using Snap for installing packages.

    Side note, if it wasnt for Snap using a proprietary backend and also depending on AppArmor (generally regarded as a weaker MAC than SELinux), I would prefer Snap over Flatpak. It creates a better sandbox (aka the actually Security of the software), avoids sandbox escapes, blacklists against broad permissions (e.g. $HOME access), and Snap packages generally have stricter permissions (which determines the real-world security of Snap). Sandboxing is very important for Desktop (and server) security. Android is does the best job of this, but it would be nice if projects like Sydbox, Crablock, or Bubblejail were adopted and built-in to the package manager.

    But even without any of the previously mentioned problems, I just think Fedora is a better OS. Fedora comes preconfigured with SELinux policies to confine system services they are quicker to adopt new technologies. Fedora is also a semi-rolling distro, meaning packages are quicker to get updated than on Ubuntu. Fedora stays FOSS, where as Ubuntu becomes more locked down. Also, the package Brace made by the developer of DivestOS is great for quickly hardening a Fedora system.