You mean they asked Claude to rewrite their app in Rust like fucking everyone else is doing rn?

I used to run dockprom but it was pretty ressource intensive and I have since moved to Beszel + Uptime Kuma.

I have yet to find a way to keep logs between container re-creation (I unsuccessfully tried to set up Loki back when I ran dockprom), both Beszel and Komodo only display logs from the container currently running in the stack.

I have no idead if or where LLM’s are being used

Having read previous Bluefin announcements, the Bluefin maintainers have fully embraced LLMs, so I wouldn’t be surprised if they made extensive use of them for Bluefin Server.

Personally, I would put one of the 990 in the server, use it as system drive + storage for databases and such. Then, the EVO in NAS to store backups, as it’s not very bandwidth sensitive and saves a little bit of lifespan out of the main SSD. And probably an extra copy of your restic repository onto the external 2TB SSD (though I would store that extra backup offsite on Backblaze B2)

Oh, and you should probably set up restic’s rest-server, it makes backuping from other computers really easy (easier than SFTP, which I had issues with)

and probably everything else too

*Takes a look at the first three folders*

• .agents/skills
• .claude/skills
• .cursor/rules

Yep…

After a couple of years using Kagi, I’ve been giving Uruky a try for the past two weeks. It’s a European meta-search engines which uses European indexes.

In my experience, setting Serper (a UK-based privacy preserving proxy to Google Search) as the first index to use gives the best responses, but I still happen to go back to Kagi once in a while (like, maybe 2-3 times in the past week).

I also have it, I find it quite tricky to work in it because of its compactness. In retrospect, I would have gone with a Fractal Design Define and some hack to enable drives hot swap.

Honestly, posts and users don’t take that much space on disk, especially relative to images. I rarely bother to clean abandoned communities on my instance.

What is the use-case? If it is for your monitor, you would be better served by purchasing a second-hand color calibration probe and use DisplayCal, especially as screens’ colours tend to shift, you should do it regularly.

No, you couldn’t because the PS3 has a very different architecture compared to most PCs. That’s what projects like RPCS3 have struggled since the beginning.

Someone at lemmy.world has to do something, it is the third or fourth account of this type in less than 48 hours.

Honestly, it makes sense that Lemmy refuses to use SVG as profile picture, as they can be used for XSS.

The real issue here is that dd.mm.yyyy can be confused with mm.dd.yyyy (I do not thank you, damn Americans), whereas yyyy-mm-dd is unambiguous.

Feels like it should be on !nottheonion@lemmy.world

I do not believe Lemmy has a migration mechanism in place like Mastodon for example. Unfortunately, you will have to re-create both users and communities on another instance.

It has only been available for 2h30 on NPM, so unless you had the misfortune of installing the latest version in this short window, you should be fine. Thankfully people have been able to quickly catch this.

Yes, but usually when you use automerge you should have set up a CI to make sure new versions don’t break your software or deployment. How are you supposed to do that in a self-hosting environment?

I guess auto merge isn’t enabled, since there’s no way to check if an update doesn’t break your deployment beforehand, am I right?

I learned yesterday that Codeberg is only free for open-source projects, not closed-source. I believe there are other Forgejo instances that accept closed-source projects though

I’m thinking of using Dockcheck. It’s not a drop-in replacement for Watchtower, but you probably can wip up a quick systemd service to run it.