Apologies for the length of this post. I did my best to make it succinct, however it's still too long for Discord, so the rest will be put in the first post.
Some background on me:
-
A long-time GrapheneOS user, I noticed similarities between GOS and Secureblue, and I'm trying to avoid repeating early mistakes that compromised security unnecessarily.
-
I've read Secureblue's documentation and this linked blog post from the FAQ (https://madaidans-insecurities.github.io/linux.html), and if these issues persist despite Secureblue's best efforts, I'm open to using Windows if it is a better option for me. Ideally this would not be the case.
-
I also reviewed Qubes discussions on this server, but its reputation here as a "reasonably secure" project is unclear, leaving me uncertain about its suitability for my needs.
Here's my priority list, in order of importance. I'm trying to balance security needs with a limited budget that will remain constrained for a few years, limiting my ability to isolate tasks with separate devices:
-
Browsing the internet: I'm on board with hardened Chromium and official Tor Flatpak, and open to Mullvad Browser if recommended.
-
Local AI usage: I want to use AI for rewording online posts to avoid stylometry analysis (though I'm unsure about its effectiveness), keep my writing succinct, and as a starting point for research, with the following requirements:
• Sandboxing with no internet access
• I have a 3060TI, but Secureblue's resource usage is a concern
• I'd like to explore any options for allowing extra leniency within a sandboxed environment, possibly by adjusting settings with ujust or finding an alternative solution
(cont'd below)