Uh, is this closed source?

I’m in favor of heavy AI users adding an [AI] tag to their posts. I don’t think we have to worry too much about AI users trying to pass their work as organic. Almost all AI users I know proudly and annoying shout about how great the AI is. So I think only a minority of AI users would try to hide their AI usage.

I think we should encourage users to add [AI], but also make it voluntary. We need people to complain in the post when someone doesn’t add the [AI] tag, but they should have. I’m guessing here, but I think the more people complain about the missing [AI] tag, the more likely the project used more AI. I’m guessing if the AI usage is low, like 1 or 2 commits out of 1000, then not as many people would complain.

people like myself who love to work on projects, but often shy away from the commitment, discipline, and responsibility of seeing them through

Username checks out.

You don’t like the one you posted to the Jellyfin comm

I don’t know enough about it to form an opinion about it.

If not then why share it?

Because I thought it was interesting.

If you want GNOME: Bluefin or Bazzite.

Why Bluefin or Bazzite over just regular Silverblue? I’m running Silverblue on a Thinkpad and all the hardware works fine.

I installed Fedora Silverblue on my parent’s laptop almost a year ago and I haven’t had any complaints or issues.

They’re really not tech literate or heavy users so Silverblue is the perfect fit. I installed and configured Librewolf and Bitwarden for them and everything has been running fine. Everything else is vanilla Silverblue.

They don’t know or don’t care about updating software. But Silverblue does flatpak updates automatically in the background. OS and firmware updates are integrated and handled via the Gnome Software Center, so I’ll click the install button every so often when I visit. No terminal required! There is a password prompt, but at least it’s a GNOME shell password prompt, not a terminal password prompt.

Additionally, I was able to get LUKS encryption working without my parents noticing: https://discussion.fedoraproject.org/t/encryption-advice-for-silverblue/162810/7

It’s not the most secure LUKS implementation, but I’m also not worried about state actors hacking my parent’s laptop. Originally, I skipped the disk encryption entirely because the extra password prompt made it harder to use the computer.

Update: Actually, maybe there aren’t any password prompts to update the system. Last night when I shutdown my laptop, I saw there was a “critical update” in GNOME Software Center. I left it alone and didn’t click anything. But when I went to shutdown the laptop, the GNOME shutdown dialog had a marked checkbox that said something like “also apply OS updates on shutdown”. I clicked shutdown and (again no password prompt) the computer shutdown. When I rebooted the computer today, I see there are no more pending software updates!

Streaming server: Navidrome

Desktop client: Navidrome web

Android client: Symfonium

Nice! I’m very tempted… I really like that it’s SailfishOS. I bought an Xperia 10 III with SailfishOS and it works pretty decently. I’ve used the Android compatibility layer, it’s pretty good.

I know it’s not 100% FOSS, but maybe it’s seeming like we could rally around SailfishOS for an alternative to Android and iOS.

Android client here: https://github.com/fluxerapp/flutter_client

Does this actually work though? Seems like it would have the same problems that the AI essay detectors have.

¿Por qué no los dos? I own an EV for city driving, but I also own a gas car for road trips.

having to spend an extra 30 minutes charging an EV on a road trip

I’ve taken the EV on road trips a hand full of times and sometimes it was just 30 minutes. Other times, it was waaay longer than that because there were 5 charging stations, but, surprise, 2 of them are broken and the rest have long lines of cars waiting.

Personally, I find those situations pretty stressful, so I don’t take the EV for road trips now.

This is an area where we want to be honest: we don’t have clean answers, and we are not going to pretend otherwise.

AI tools including language models, image generators, and code assistants, are genuinely useful. They can help someone start a project they wouldn’t have had the skills to begin otherwise, prototype ideas quickly, and finish projects that might otherwise be abandoned. Even the UnfinishedProjects platform likely wouldn’t have come to fruition without the assistance of LLMs.

At the same time, the ethical picture is genuinely complicated. Large AI models have been trained on vast amounts of creative work scraped from the internet, including work released under open licenses, without meaningful consent from the people who made it. The legal landscape regarding AI is unresolved and messy at the time of writing. In addition, there are broader ethical concerns regarding the impact of data centers on local communities, the environment, and the economy. In short, the ethical considerations are vast and widely recognized.

The current landscape has left the broader community divided on the ethical and legal implications of using AI in their work. There is no single policy that we can implement on our platform to align with everyone’s views, and we believe the best way forward is to allow each project to determine its own boundaries regarding AI.

Individual projects set their own AI policies. We do not mandate a platform-wide rule on AI use, because the right answer genuinely varies by project type, contributor values, and context. Every project is encouraged to state its policy explicitly using the Template:AIBanner at the top of its project page. By being direct and upfront about AI use, you provide transparency for each contributor, allowing them to make an informed decision about whether to participate right from the start, rather than uncovering that information later.

The quality bar does not lower just because a tool helped. In our view, pasting unreviewed and/or an untested output is not a contribution. Reviewing, editing, testing, and taking responsibility for the work — that is the contribution. The tool is just the tool.

We expect this area to keep evolving, and our thinking will evolve with it. As with the rest of our wiki, we expect these pages to grow as members bring more insights to the table and as the technological landscape shifts.

dick move.

Which users are doing that so I can block them?

Also it’s reasonable to think discussion forums are in some sense ephemeral

This is 100% wrong. This isn’t Discord or chat. People expect forums to appear in online search results, i.e. be persistent.

CEO later added: the exec team is still getting a raise though.

Does UnfinishedProjects have a stance on AI? Are AI tutorials OK to post? Is AI slopware OK to post? Are AI answers to skill exchange posts OK to post?

Might be worth mentioning that somewhere on the site.

What’s up on piefed.zip?

GitHub issue about this: https://github.com/jqwik-team/jqwik/issues/708#issuecomment-4554650392

the agent detected and refused the injection on first contact

Shame. Prompt needs more work.

Maybe instead of deleting the code, it should do something more subtle… like telling the agent to generate (even more) mountains of code and introduce subtle bugs, crashes, and sleeps.

Most native apps collect far more data than their website equivalents ever could. They request permissions to hardware, sensors, and background processes that browsers deliberately restrict.

On March 27, 2026, the Trump administration released an official White House app for iOS and Android. … Apple requires apps to submit a privacy manifest disclosing what data they collect. The White House app declared an empty array. Zero data collection. Meanwhile, the actual binary contained ten analytics frameworks, including the full OneSignal SDK with a sub-framework specifically for location tracking

Hm. Didn’t think about it like that.