More engaging posts also means higher quality.

A quality post don’t let me guess what Hollo even is, but use a few words to describe.

Just posting links kills engagement.

Really

Never ever heard of

Thanks for quality post

(Luckly immich is much better quality than this post)

You actually WANT to be with low free memory. Provided that most of it is used by cache.

Free memory is a waste, when you could cache stuff for faster access.

That’s how Linux memory management works, and it make sense if you relflect on it. Better cache that page or that file that is used often, since free memory is just wasted. Cache can be freed and memory reclaimed in a fraction of a millisecond when needed.

So don’t bother too much. Unless your SWAP usage is high, don’t bother.

Also consider that Linux kernel will use your swap a bit even if you have lots of cache, because the kernel knows better than you how to improve your performances. Swapping out never used stuff is better than killing cached items.

Again, don’t oberthink memory on Linux, the best alarm is when swap is constantly happening, then yes you need more ram (or to kill that broken process that keeps hogging due to a bug)

You don’t get it: I am and will remain the only user of my instance…

Do you even now how Lemmy works? Did I say I was going to let ICE people create users on my instance? I only said I don’t defederate any instance.

Wake up. Ice, being government, it’s already legitimized enough in real life.

What difference would it make in the social media. Better if they are out in the open in social media instead, at least they get responsible for what they post, officially.

I don’t ban anyone or any instance in my own instance, so no they cannot be “parammanned” from Lemmy. That’s not how it works and why i like Lemmy and its principles.

(Not American here)

While i agree fediverse is then solution and i don’t use bluesky, i don’t see the issue is recognizing ICE as verified.

After all ice is a government agency of the USA whether you like it or not, and should be verified if there is a procedure to do so.

No i don’t like ice and i do not condone what they do, but that doesn’t change the above statement.

Really? Wow… I assumes was USA LoL

Well, I think there are a few… Indeed I know at least two italian instances, plus my private instance…

But most of activity is currently on USA instances like world, so…

My recent experience regarding questions on documentation:

• dovecot: shitty useless responses, totally made up
• Gentoo linux: to be checked twice and mostly wrong or fake
• godot: accurate and correct almost always, maybe examples not always 100% correct
• C++ standard 17: correct, never had a wrong reply from llm, also the exact ples where on point and correct

I think that’s all what I have used it for in the last six months.

Note: I used only Google search AI llm, nothing else.

So it seems that depends on what you ask.

Maybe you should check 5hat again. I never used llm before 2025, but proved itself useful for a few tasks. Yes check and verification still needed, but indeed made my life easier and got taken done faster. Quality was still a good as what I could do myself. Maybe that doesn’t speak well of myself I don’t know.

While you are correct, as all tools AI is not bad per se.

If you use ai to replace more lengthy documentation searches and write your own code that works out pretty well and speed up your work without degrading your coding. Granted, I got plainly incorrect answers as well, but at least I managed to be much more efficient.

Treat LLMs/ai as a glorified documentation aggregator and that’s how you correctly use that tool.

Like, use a knife to cut and cook meat, not to cut another person body, and that’s how you correctly use that tool too.

I run a setup very similar for many years. Upgraded progressively from 2x120GB to 4x4Tb, from mechanicals to sdds.

I can say don’t cheap out on the usb enclosures, the more you pay the better it is. I purchased a 4 bay JBOD usb3 box with a fan (150€+ nowadays) and that is the only enclosure that really worked out and still works (but retired) today. All single disk enclosures will fail sooner or later depending on how cheap they are, just take that onto consideration.

The setup itself is pretty good and stable, I would suggest standard Linux MDRaid 1, and on top of that something simple like ext4. I wouldn’t put anything that adds to the disk workloads like zfs, but maybe I am wrong.

Speed wise, I was able to stream movies without any hiccups, and that’s plenty I think.

Do not cheap out on the enclosures. Cheap ones will last 1 month, I don’t kid you.

And keep them cool… Fan… Air circulation… USB controllers will be killed faster than mechanical disks by heat. And 24/7 will generate heat… Those enclosures are not built for that…

Again, it’s pretty doable and I did it for almost 2 decades. DONT BE CHEAP ON ENCLOSURES (did I say so already?) And you should be fine.

If you don’t want to use postgresql for some obscure reason use MariaDB real open source MySQL drop in replacement.

This.

That’s why temping obscurity for security is not a good idea. Doesn’t take much to be “safe”, at least reasonably safe. But that not much its good practice to be done :)

My server has 48gb ram and in top Lemmy doesn’t appear even in the 0.1% memory usage.

Disk space 10gb, CPU/ram not noticeable on my server (lots of other services using more than Lemmy).

I think it’s been up about one year. One user but I subscribe to all communities I find remotely interesting.

Running “my” own single user instance here.

Great! Love it! The whole idea.

Agree on the handpicking yes. Nothing bad in reposting content, but mass posting just drowns communities and doesn’t generate engagement.

No thank you. Just leeching content from reddit will never get Lemmy anywhere.

I blocked all content reposted from Lemmy (there was an instance dedicated to that iirc).

First of all, posts without comments are pointless. Interaction and comments is what grows Lemmy. Second, bot posts are not well received by most here I think.

I can see it useful only as a “reddit backup” of kind, in case the platform goes offline for good and all that is lost.

I would worry for the level of moderation required to filter shitty / ai / bots etc.