fyi: some fellow forked all of the sentry stuff from when it was bsd and made it less shit to run yourself. I’d been running sentry as self hosted since it came out effectively and they made some real questionable outright bad architecture decisions over the years that made running sentry super annoying to get you to use the saas product. Like yes I want to use Cassandra and elastic search for my one user error logging

regardless https://glitchtip.com/

So this is literally a rehash of Spyce, a CambridgeMA based startup whos first restaurant was in the old chipotle in harvard square. They had a similar setup but one that was able to make actually-hot food instead of salad bowls which were decent. (they ironically sold to sweetgreen in 2021)

I can speak for the tangara, lovely hardware but it is an mp3 player first and foremost. the hardware is technically capable of network stuff but their software stack is mostly focused on recreating that iPod experience and keeping the battery life to days of playback. It does take some getting used to if you’ve been living in musical algoworld for a minute

there’s been something that’s really been rubbed me the wrong way about jeff in the last few years. he was annoying before and had some insights but lately I’ve been using him as a sort of a jim crameresque tech-take-barometer.

What really soured me was after he started picking fights with some python people a few years back because someone dared post that a web framework? (couldn’t dig up the link) was a greater contribution to the world than S/O? His response was pretty horrid to the point where various python leaders were telling to stop being a massive dick because he was trying to be a bully with this “do you know who I am” attitude because he personally had not heard of the framework so it wasn’t acshually at all that relevant compared to S/O.

and now this combined with his stupid teehee I am giving away my wealth guise look how altruistic I am really is a bit eugh

in my headcanon it is and he’s just been moonlighting tv for the last 25y

Your phrasing of the question implies a poor understanding. There’s nothing preventing you from running containers on bare metal.

My colo setup is a mix of classical and podman systemd units running on bare metal, combined with a little nginx for the domain and tls termination.

I think you’re actually asking why folks would use bare metal instead of cloud and here’s the truth. You’re paying for that resiliency even if you don’t need it which means that renting the cloud stuff is incredibly expensive. Most people can probably get away with a$10 vps, but the aws meme of needing 5 app servers, an rds and a load balancer to run WordPress has rotted people. My server that I paid a few grand for on eBay would cost me about as much monthly to rent from aws. I’ve stuffed it full of flash with enough redundancy to lose half of it before going into colo for replacement. I paid a bit upfront but I am set on capacity for another half decade plus, my costs are otherwise fixed.

I assume ppl still run bzflag servers

The way I’ve been using it for a few years is that most of my machines can see each other and I have a shared folder and versioning setup. As I add things they move between the different machines and once an additional machine has it it is available to the others until everything is in sync

You can definitely do chain topologies which are useful for certain things with a single source of truth

As a note, I believe that syncthing will actually scale up with more nodes as they will all share with each other if they know each other. If you’re doing this 1 to many then this is not the case of course.

Seeing as the XLibre fellows upstream commits were reverted because they were absolute dogshit, I don’t think that the fork has legs

Sidenote: I almost ended up working for the company almost a decade ago now lmao. The board was full of other characters we all know and love here. The €€€ offer was high for EU, but I still laugh at the growth potential of my 10000 dollars yearly equivalent in their tokens. The website was unique in that it scrolled… up. I think it’s still on archive dot org

I wonder why his 10000 agents haven’t done the work yet. It seems like such a straightforward plan.

Ive actually been personally moving away from kubernetes for this kind of deployment and I am a big fan of using ansible to deploy containers using podman systemd units, you have a series of systemd .container files like the one below

[Unit]
Description=Loki

[Container]
Image=docker.io/grafana/loki:3.4.1

# Use volume and network defined below
Volume=/mnt/loki-config:/mnt/config
Volume=loki-tmp:/tmp/loki
PublishPort=3100:3100
AutoUpdate=registry

[Service]
Restart=always
TimeoutStartSec=900

[Install]
# Start by default on boot
WantedBy=multi-user.target default.target

You use ansible to write these into your /etc/containers/systemd/ folder. Example the file above gets written as /etc/containers/systemd/loki.container.

Your ansible script will then call systemctl daemon-reload and then you can systemctl start loki to finish the example

The Kde store UI is a bit lacking the screenshot browser has arrows all the way at the edges.

I uploaded three screens two of the configuration and one of the widget on a panel, expanded with all the advanced controls.

Edit: Ive reuploaded the images with the expanded widget first

During my expirementation with some of these self hosted llms, I was attempting some jailbreaks and other things and thought would this be any good at ERP?

Only if youve never been with another human being.

The fork Ansel, is supposed to improve on the UI situation.

Unironically their greatest movie.

Kate has excellent lsp support nowadays as well.

Pretty solid gameplay so far, threw a tenner at them for the work

I think the gap you have is in understanding that Podman Compose was meant to line up with the limitations of docker’s compose, but technically is more capable.

Quadlet files let you do more complex workflows like deploying multiple copies of a service in your deployment that regular compose doesn’t, while not running full kube.

The use I have is that I have something deployed in compose right now that I’d like to scale up on the box since i have the capacity for it, but dont want to deal with a full kube setup or the politic

Personally I’ve converted most of my single node k3s to using quadlet files instead as its less fragile. I absolutely deploy single containers in the quadlet. They show up in journalctl and the ergonomics are great.