Update I have come to a decision. Thank you to all who contributed suggestions. Please feel free to keep the discussion going to help others.
I only mentioned physical port attacks in a much larger list of things Linux MUST improve on. I am not a grapheneOS shill, nor did any of the supporting articles I sent relate to GOS, so I don’t really understand your response. Read through the links I posted and learn more about the operating system you use. I am NOT saying linux is dogshit, I very much love linux. Why not just educate yourself on this topic instead of assuming things from a place of ignorance or constructing a strawman. I spend multiple hours per day reading and putting into practice Linux hardening techniques, I am not just working with a surface level understanding of Linux security.
Even open source is vulnerable. Two questions: do you examine all the commits on every app you use? Do you compile every update to the apps you use from source? Sandboxing is important because if an application is compromised it cant lead to privilege escalation or userspace spyware.
I‘m not that bad at rhetoric either but I avoid it when I can.
Your argument is empty. Privilege escalation attacks are plain old cves that get found, evaluated and fixed. You need access to the phone, mostly in an unlocked state to get anything to work like that, same as with a computer.
I know a couple of pen testers and I would definitely know if there were large differences between operating systems securitywise.
CVEs are often go mislabeled as normal bugs and dont get the attention needed. It also may take a bit for such vulnerabilities to make it downstream.
A simple privilege escalation attack on basically every system goes as follows: add a function into the bashrc file of a users that runs a script, have the script intercept the users sudo credentials and pass the command on normally as if it was just the regular sudo command. Now you have root. Nothing here requires priveleges beforehand. Anything, be it a script, appimage, malicious binary, etc can follow those steps and gain root access by compromising the wheel user. Even without compromising a user, it could simply add a Systemd user service that keylogs (keylogging is still possible on Wayland without security hardening)
A prerequisite of course is getting that file onto the user’s computer. There are a plethora of ways. Simplest way is to learn what applications the user installs, find the weakest link, and compromise them.
There are of course much more sophisticated and better ways, some of which are detailed in the supporting links I sent. Every Security expert and researcher I have talked to can recognize that Linux has an outdated security model. The best links to read would be the hardening guide and “linux isnt secure”.
I did quite some reading in my time, as I mentioned. The methods you are describing are riddled with ifs and buts. The reality is that even online systems arent hacked if they dont have obvious flaws like passwords in root ssh. on the other hand tools like john the ripper can break each and every common encryption given the right circumstances. Its no difference. Its all just marketing.
