

Minimalism. Compared to AOSP, Google components and pings removed. Compared to other privacy GSI ROMs, no weird, quirky, or flashy functions or themes the author decided to bake in.


Minimalism. Compared to AOSP, Google components and pings removed. Compared to other privacy GSI ROMs, no weird, quirky, or flashy functions or themes the author decided to bake in.
Whatever the choice, there will be trade-offs. From a technical standpoint, there shouldn’t be anything wrong with the upcoming Motorola phone with GrapheneOS as long as the bootloader is still unlockable and relockable so that the user can install a known good copy of GOS instead of trusting what Motorola put on it.
This is amazing and stupid easy indeed. I already have to manually download yt-dlp periodically (Debian btw), might just switch to this instead.
I always forget the extensive options and syntax for yt-dlp and this is perfect for when I don’t quite care enough about some videos to download in the default highest quality or if I don’t care about the video part at all.
As much as I trust them with passwords. Which is not too much trust. Implementations of passkeys also tend to be frustratingly bad.
As long as you have a strong backup strategy, I would recommend full disk encryption during installation, especially if for a laptop. Peace of mind with negligible cost on modern hardware. Even accessing the encrypted disk from a live USB takes only two extra commands compared to an unencrypted disk. As long as the LUKS header doesn’t corrupt, hence the need for good backups.


This is one of the mundane scripting tasks I would have my local LLM help with. I would still read through and make sure I understand the resulting script before running it.
Wanted to, but lacked the motivation to learn it. Was stuck on one occasion without nano, so I pulled up the vim cheat sheet on my phone.


Not going after a bug though, it’s just the way the included battery meter in Xfce (and other X11 battery indicators I’ve found) works, while things like Android track usage over time to give a better estimate.
But based on the other responses, it looks like I’ll have to cook it up myself.


As a daily LibreOffice user, I agree with you on the UI. I can’t even keep track of how many different settings menus there are and each of them are a labyrinth unto themselves. What ended up saving my sanity was setting the UI to single toolbar and purging every unnecessary button in Calc and Writer. Might be unpopular, but I then arranged the remaining toolbar features the way they do in Google Docs. For Impress, I set it to the tabbed ribbon-esque interface.


Not that I’m aware of


That is correct, IIRC, the mismatch does limit how much of it can run in dual-channel. Even if a single stick is natively 24 or 48 GB, there is additional strain on the memory controller. It is the way it is on my setup since I had planned an upgrade to a full 64 GB and was holding off until a good deal on the remaining 32 GB kit, which will never come unless the AI bubble bursts.


I think I could hack such a feature into the tray indicator as a weekend project, but wanted to see if someone already accomplished it before I go reinventing the wheel.


What about it? I see it kicking in at least 10 GB before my RAM is full and I haven’t noticed any fundamental differences between how zram works on my 48 GB workstation and my 8 GB devices. Maybe I’ve never had a workload that filled all 48 GB + extra zram capacity, but it’s never given me an issue.


My workstation has 48 GB RAM with 50% allocation allowed to zram, no disk swapping. It works just fine. Once I use up the majority of my RAM, it kicks in the same way it would on any other system with less RAM.


Pro:
Con:


That really stinks. Does the audio version do anything different?


16 GB VRAM GPU, models stored on SSD, rest of the computer doesn’t have to be crazy. Intel Arc is best bang for the buck at the moment. You can get LLM running on 8 GB cards or even the CPU, but IMO such small models are more novelties than workhorses. I personally use Debian but you’ll be fine as long as your distro’s repo has drivers recent enough for your GPU.
For perspective, I’m using such a build to help with boilerplate code, single-use scripts that I don’t have the patience to trial-and-error (like ones that have to deal with directory structures and special characters), getting an idea of what’s what when decompiling and reverse engineering, brainstorming tip-of-the-tongue ideas, and upscaling images.
I was so excited I could finally solve it, but unfortunately, it already was set to Latency.
Nearly everything that both requires a phone and disrespects my privacy has been work-related, so using 2 phones has been a solid choice for me.
The work phone has a sim from a mainstream carrier and only gets powered on while at work during work hours. Maybe I’m spoiled that my workplace tolerates this arrangement. I couldn’t imagine having to be reachable any time of the day. I didn’t intentionally buy a separate phone, it’s just my old phone repurposed.
The personal phone has an “IoT” SIM which can be purchased non-KYC where I live. All FOSS apps and a personal number via VoIP.
I know it isn’t by any means airtight, but it gives infinitely more peace of mind than just trusting whatever sandboxing mechanism available on one device will be sufficient.