The initiative is great, hands-down. But… Is it only me, or it’s somewhat strange? The biggest prominent feature on the website is a video that doesn’t seem to be voiced by a real person. FSFE that supposedly backs the project in terms of reputation… what is it? I’m aware of FSF (Free Software Foundation, somewhat US-aligned) but not of FSFE. Reading about it from wikipedia, it seems legit. Did some people known in the Open Source community endorse them as well? I’m essentially willing to build my chain of trust to them, if it works.

Sometimes it’s fun to smash some AI in the game Feudal Tactics (or get wiped out). Not a thorough game, but it’s surprisingly repeatable if you’ll get along with its vibe.

And generally F-Droid.

What is the AI isage disclosure for this project?

Indeed. For those who are relatively new to IT: it’s a reference Embrace, Extend and Extinguish, a strategy known and legally confirmed to be used by Microsoft.

Personally, I struggled somewhat until I’ve found the Search and Run a Command action. It’s bound to / on the keyboard by default I think. Otherwise accessible via Menu > Help > Search and Run a Command.

So if I’m clueless where a certain action is in the menu, I just press /, then type my action as I think it could be named, then try it out. This, together with simply my memory of how to do things, covers 97% of my needs. One time out of 20 or 50 I encounter something that I don’t know, at which point I duckduckgo, and the experience can be much worse.

Would you say this flow sucks? Or actually, it’s fine? It’s up to you to decide that.

Personally, I exercise the “growth mentality” here. Meaning, I don’t compare with a target value such as “I’ve converted 10% of the Earth population to OSM”. Instead, I compare with OSM as it was 6 months ago. This approach gives peace of mind and steady progress for me personally.

I’m a sporadic/infrequent openstreetmap contributor, too - although I’m not second-guessing these contributions. Useless or not, I do use OSM all the time personally, so reason stands somebody before me did a good job and helped me a lot.

Admittedly, some of the people who use OSM around me use it because at a certain moment I’ve explained them the advantages. (I live in Europe and people frequently use other modes of transport besides the car, so OSM is legit frequently the better choice.)

To be a bit nitpicky, Tesla has a load of shit reputation besides having a con artist and fascist CEO. Worker’s rights abuse, accidents, false claims (lies-lies-lies).

Though it is true that Elon is *so* shit that buying Tesla would be no-go and shameful on my book even if the above wouldn’t be true. So you have a point. Personally, I categorize it as Mullvad’s founder being misled rather than genuinely corrupt and evil. But it’s subjective, and I see and respect if people think differently.

It is noteworthy that Mullvad did not donate any money to any political party.

is solid for free

Just a heads-up, keep in mind the community you’re posting in (Open-Source). Open-source is about the open-source development model, not whether you’re required to pay.

Sounds interesting! And welcome to Lemmy if you’re a new user indeed! (Otherwise, cheers.) Few personal thoughts below.

Atomically upgradable images and the added security sound interesting to me. I’m not specialized enough to say whether that’s really going to help for my threat model, personally. I mean, in /boot I currently only have the Linux images. Verifying those seems sufficient, because the actual system is only decrypted afterwards when I enter my password.

I must also say, accepting a concept is harder for me if the proposed line of thought is “GNOME OS is to be loved, and %project_name% is a bit like that”. Because personally, I remember Gnome since version 2 and I dislike what they’ve done with the system since. I only use gnome-disks from the entirety of them. From KDE I use more (kate), and from XFCE I use way way more. And I also dislike how GNOME pushed systemd onto all of us as a hard requirement. Not that systemd is necessarily bad, but the push for it didn’t feel right.

Yup, that’s simply a deal breaker. [For me].

If it’s for you, too, you could maybe even mark the post with [SOLVED] or “Update:…”. If you wish. No obligation.

“Frame” is AI slop. I think it was explained that the whole or most of the project is written by Claude Code. https://github.com/isene/frame#license

How many bugs there are and for how long it’s going to be maintained, anyone can guess themselves.

Interesting, checked it now. Phoenix: https://git.dec05eba.com/phoenix/tree/README.md

It’s Wayland though (and partially Xorg)? I mean, the protocol that the client has to speak with gamescope is wayland (or, somewhat indirectly, Xorg).

I’m curious about your statement … (on the bugs in kanata)

I’m on Linux, and I get hit by this bug a lot: https://github.com/jtroo/kanata/discussions/422

The official work-around is to use an additional F24 key press-release sequence: https://jtroo.github.io/config.html#tap-hold but that works poorly for me. Some apps are messed up slightly because of this (editors such as kate which stops understanding Ctrl Z properly), some work smooth, and some even break entirely to the point of becoming unusable (terminals such as raw tty3 or xfce4-terminal).

I’m curious about your statement

I’ve just recently installed GrapheneOS on a new Pixel 8 (as in, refurbished, but for me it was new). It worked quite well. I’ve used to use the CLI installer before, but after looking at the current recommendations on their website, I followed through with the web-based one. One point of local struggle for me was when I needed to “Flash factory images”. It looked like it’s not going well, I didn’t understand (still don’t) how to read the instruction unambiguously on when I should stop waiting, ended up restarting my phone once needlessly. However, doing the same step again afterwards somehow made the flashing successfully, and the button for the next stage got enabled (“locking the bootloader”). In total, it was inconsequential.

Everything is working super-smooth since. I’ve did a hardware attestation procedure to make sure that my refurbished phone is fine, it and shows up as fine. I’m quite content with how it went. Considering how you describe your phone usage, you would benefit from GOS or another open firmware a lot. You can give it a try, it’s a good firmware! I wish there’d be projects that would be freedom-first, not privacy-first with the same quality level. I digress and it’s subjective at this point. Just give it a try!

Adding to the comments about VIA and QMK, you can also use something on the computer itself to do ergonomics tricks and making F7 usable again.

I’m in particular using kanata, which then works on everything from notebook keyboards to custom ones. I’m not 100% happy with it and in fact considering to write my own software for this, but it’s good-enough for basic usage. It might be good-enough for you. kmonad is less buggy but has fewer features (definitely enough for F7 though).

Maybe some AI crawlers started to honestly report Linux instead of Windows as they possibly did before? Otherwise such a quick spike is hard to explain (unbelievable). Or did I miss any big news lately?

//I’m on Linux in the last 15 years and I don’t even care that much about what unrelated to me people use, although I always do help fellow Linux-ers when I encounter them in the wild.

Good starting point would be to focus on projects hosted on codeberg. Their official policy since recently is human-first. (The rules are detailed though, I don’t want to over-simplify. But very reasonable and vibe-coded stuff is definitely out of the window.)