The main image shows mobile devices, but I don’t see a mobile device build. Are you planning to add releases with precompiled apks?

nixos-hardware go brrr

https://xkcd.com/2501/

Did you remove the french language pack? That’s gonna be another 15 easy

I finally finished setting up my Nebula network! An overlay network, as opposed to a true VPN, but excellent for flexibility and remote access. For anyone wanting maximum control over your network with excellent performance, I highly recommend it.

Check out apalrd’s blog for a great tutorial if you’re interested.

I’ve never heard of that in the states. What region are you referring to? Sounds like an eastern seaboard thing to me.

For that workload? I quite literally run more than that on a (le)potato

Apparently I’m in the minority, but I love Logseq. I’ve used it with Syncthing for personal notes and grad school for the past three years with no hiccups. Maybe my success with it is partially due to nested bullet points already being how my brain works but the default paradigm is perfect for me.

The plain markdown files are organized reasonably, so I can straight up use Vim as my notes editor if I want.

Tags (#) create a new page to easily circle back to topics later without interrupting your thought pattern to make that structure manually. Once you leave edit mode for the line the tag becomes a link to that page. Some of my favorites are #clothes-that-fit (where I can easily embed a picture of the tag of what I’m trying on to look for deals online later), or #reading-list.

It’s just so useful.

if you could start again in your self hosting journey, what would you do differently? :)

That’s an excellent question.

If I were to start over, the first thing that I would do is start by learning the basics of networking and set up a VPN! IMO exposing services to the public internet should be considered more of an advanced level task. When you don’t know what you don’t know, it’s risky and frankly unnecessary.

The lowest barrier to entry for a personal VPN, by far, is Tailscale. Automatic internal DNS and clients for nearly any device makes finding services on a dedicated machine really, really, easy. Look into putting a tailscale client right into the compose file so you automatically get an internal DNS records for a service rather than a whole machine.

From there, play around with more ownership (work) with regard to what can touch your network. Switch from Tailscale’s “trusted” login to hosting your own Headscale instance. Add a PiHole or AdGuard exit node and set up your own internal DNS records.

Maybe even scrap the magic (someone else’s logic that may or may not be doing things you need) and go for a plain-Jane Wireguard setup.

Another +1 from me. Very similar setup and it’s been working for me for years.

So folks know this isn’t a hypothetical Edison Motors 🍁

I’m genuinely not sure what has you so reactive here, but go off queen

Typically the primary product of the farm would be the livestock though, not the manure. I’ve never heard of someone using an entire farm just for collecting and processing manure.

What exactly is meant by “manure farm” here? If they mean “farm that fertilizes with manure” then that’s a weird jeer since it’s not uncommon.

Also an 80 year old man doing this sounds hysterical.

To staffers - What do the kids like nowadays? What’s that? Go chella? Yeah give me a gochella rally next

Coarse Salt. Add just enough water to move it around + a little dish soap and shake. Works like a charm.

Even in the US, state-level representation hardly counts as local. Neighborhoods, towns, counties, etc. all have people representing them.

It’s cool to care about and build up your community.

Excellent notes. If I could add anything it would be on number 4 – just. add. imagery. For the love of your chosen deity, learn the shortcut for a screenshot on your OS. Use it like it’s astro glide and you’re trying to get a Cadillac into a dog house.

The little red circles or arrows you add in your chosen editing software will do more to convey a point than writing a paragraph on how to get to the right menu.

Believe what you will. I’m not an authority on the topic, but as a researcher in an adjacent field I have a pretty good idea. I also self host Ollama and SearXNG (a metasearch engine, to be clear, not a first party search engine) so I have some anecdotal inclinations.

Training even a teeny tiny LLM or ML model can run a typical gaming desktop at 100% for days. Sending a query to a pretrained model hardly even shows up on HTop unless it’s gigantic. Even the gigantic models only spike the CPU for a few seconds (until the query is complete). SearXNG, again anecdotally, spikes my PC about the same as Mistral in Ollama.

I would encourage you to look at more explanations like the one below. I’m not just blowing smoke, and I’m not dismissing the very real problem of massive training costs (in money, energy, and water) that you’re pointing out.

https://www.baeldung.com/cs/chatgpt-large-language-models-power-consumption

I don’t disagree, but it is useful to point out there are two truths in what you wrote.

The energy use of one person running an already trained model on their own hardware is trivial.

Even the energy use of many many people using already trained models (ChatGPT, etc) is still not the problem at hand (probably on the order of the energy usage from a typical search engine).

The energy use in training these models (the appendage measuring contest between tech giants pretending they’re on the cusp of AGI) is where the cost really ramps up.

Avoid AMD? Why do you say that?