Fedegenerate

deleted by creator

That’s me. I don’t regret not pursuing a career though because my achievements don’t make me feel accomplished. When I get a service running I mostly feel relieved. But, I want the things self hosting provides, so I do the work self hosting demands.

An electric pressure cooker makes it a mindless activity. Load it, turn it on, come back when it beeps, or tomorrow… Who cares, it’s sterile in there.

Then to reduce: turn it on, come back when it beeps. Dial in the times and electric pressure cookers make recipes ridiculously repeatable.

My only criticism is their food is fairly homogenous so I add some crunch on the side just for some heterogeneity. Insane effort to value ratio.

Solid question. Mostly curiosity, education, love of the game. I started with raspian + docker, and wanted to see if the grass was greener. I could probably achieve everything I’m doing now on Debian + Docker. If I tore it all down, I’d probably go back to Debian + Docker managed with Dockge (probably) just for the simplicity.

Automagic snapshots though, I still haven’t learned to take a back up before tinkering, I probably never will. So lxcs are set to snap shot at appropriate intervals, so if I try change a thing, fuck it up, it’s incredibly convenient for me to roll it back and try again. A solveable problem in Debian, I’m sure, but I like how Docker does it. Native WebUI is nice. I like how Proxmox handles file permissions, it’s running my raid and I just bind mount the zpool.

Dunno, my stuff works, I’m comfortable with it. There’s more for me to learn yet. But if I were of a learning mood, I’d probably experiment with Nix.

I use proxmox, all the *arrs are in one docker host running as an LXC. So the stack is: Proxmox> LXC> Docker Compose> Gluetun> *arrs.

I run it this way because ease of set up, as a normie Nix scares me in ways YAML doesn’t. It’s one big compose file, everything talks to each other using the virtual docker network. Update is one button. It just works™.

I have an Lxc running jellyfin as a front end, an Audiobookshelf Lxc, one running Seerr. A bunch of other front end service lxcs (Immich, Homarr, what have you). All the front ends have Tailscale installed for remote access.

Family and friends have a pi acting as an access point to Tailscale. Then it’s the nginx, pihole combo for nice URLs and HTTPS. Nothing is on the open web, it’s all behind tailscale.

I started on just a pi running docker running everything for my fiancée and I, then it just kinda grew. I originally just wanted to block ads.

Trash-guides is your friend for *arr set up. Dockstarter is your friend for basic installation of *arrs. Thousands of guides on YouTube to get you started.

Groyper started what he couldn’t finish and he knew it. Imagine initiating over a table and then getting yanked like a bitch, then going for the ‘hold me back’ to try and save face. We saw you, we saw what you’re not, should have stayed sat down chatting your shit.

Why are they so fucking weird?

I understand the point is to price people out of personal computing, forcing them to use thin clients to rent compute. That doesn’t mean I’m not priced out of personal computing though.

I’d like to make a PS3/xb360 emulating machine, and boy is that an expensive toy.

Brandon Sanderson says the same for readers and writing. Readers are great at identifying something wrong, but their suggested fixes are truly awful.

Is there a writing circle equivalent for game dev?

The beer bottle for our beetle brains. These beautiful bots bemusing these bewildered boomers don’t exist.

Old workstations perhaps? Hp z440esq things? An example of such a project.

Perhaps even that is overkill though.

Your Spotify

Shame there isn’t a “your deezer” that sounds really cool. Anyways user facing service LXCs (eg. Jellyfin) have Tailscale installed. Here’s my services by box:

I have a spare no name n100 box that I would like to turn into a network box one day, then I can manage my Airvpn and Tailscale access through that. But for now the ISP router is fine, I guess.

“Speedy thing goes in, speedy thing comes out” as the cube isn’t a speedy thing (no speed lines) going in, it won’t be a speedy thing coming out.

I used to giggle at the riders with their snoods in the middle of summer.

“Who needs a scarf in summer” I’d chortle.

Then I hit a bee with my throat at, we’ll say the speed limit. Now I ride with a snood even in the middle of summer. I’m ‘vizor up’ till at least 50mph, still haven’t learned that lesson.

Oh I know, but progressives were anti-genocide too. It’s the more centrist persuasion, to whom genocide wasn’t a deal breaker, I’m speaking about.

I reject it’s just republicans that will excuse this, there’s plenty of people in this very thread. One even said the quiet bit out loud: even if Platner raped her, it wouldn’t change a thing for them.

deleted by creator

To the people that excused, and went to bat for, a genocide apologist… What’s a little rape? Honestly, I’m surprised so many are dropping him for this, maybe there’s hope yet.

I’m sure there’s something to be said about the perceived victims of the genocide, and the rape, but I don’t have time to think about it.

Start like a butterfly landing on a fragile flower… End like a bulldog eating oatmeal. Best advice I ever got, after spelling the ABCs

You can tell they don’t understand capitalism very well, by their support of capitalism

There’s a joke I’m riffing off, that Christians read the bible to reinforce their belief in God. What do atheists read to reinforce their lack of belief? The same.

Sum total of my hardware:

Ugreen dxp 4800, the Pentium one. 32gb ram. My main box: jellyfin, arrs, immich, pihole, nginx, etc… It can’t go down. I don’t think I want an LLM here.

Beelink n100 16gb ram. Local back up, redundant pihole, immich machine learning… Generally under utilised, I’d like to move some services around, does proxmox have an auto balancer?!

Spare no name n100. 8, or 16gb, I can’t remember. Abandoned box, It’s what I would put the LLM on, I did have it reserved for a remote back up.

I think it would need a ram upgrade, see corporate AI pricing me out of personal computing. Currently Amazon has a Crucial 32gb ddr5 sodimm module for £280. Which is too high a price for what I’d use it for.

Oh, and I have an abandoned gaming rig with a gtx970, and some rPi0/3s

I’ve put Ollama on an n100 before. It obviously ate all of that box and made everything on it chug, and it was too slow for human use. But if it’s just generating logs, and resetting containers then I wouldn’t mind how slow it is.

I’ve been dreaming of local AI for a hot minute. Sadly corporate AI has priced me out of personal computing, and current hardware isn’t up to it.

Maybe my n100 could, I don’t really care how slow it’s generating the tokens if it’s just resetting containers and logging errors.

Does your setup handle automagic updates, how do you handle prompt injection if so? Or, just error correction/logging?