RadDevon

joined 2 years ago
[–] RadDevon@lemmy.zip 0 points 1 week ago

Seconding Quartz, but I also wanted to share how I self-host all the services required. I self-host Forgejo and use obsidian-git to make it easier to push my vault changes to my self-hosted repo. The repo has a workflow that runs the Quartz build on any new commit, building to a location my web server points to.

It works pretty well. About 2 minutes after a change is pushed, the build with those changes is done and being served.

[–] RadDevon@lemmy.zip 1 points 2 months ago (1 children)

Thanks for the clarification. By "persist across restarts," I'm referring to the fact that if I just install the agent in my container, it won't persist if I restart the container, unless I install it on a volume which seems clunky. Running the agent alongside in a separate container with network access is the solution I was looking for.

On the Redis and Valkey restores, that makes sense. Disaster recovery is my use case anyway. Do you document the manual restore process for those? I didn't notice it in a brief review of the docs, but I may have overlooked it.

[–] RadDevon@lemmy.zip 3 points 2 months ago (3 children)

This looks exciting! A couple of questions:

  1. How would I deploy an agent into an existing docker container in a way that makes it persist across restarts? Most of my databases are running in containers.
  2. Regarding redis and valkey: what good does backup do without restore? Not trying to denigrate; I just really want to understand how that is useful.

Thanks for building this!

[–] RadDevon@lemmy.zip 3 points 2 months ago

This has been my experience with Matrix, and the message decryption problems are a dealbreaker. I hope the person who replied to you saying those have very recently been fixed is correct, but the fact that such a fundamental feature was broken for so long leaves me with little confidence in Matrix. I had this problem years ago on a Matrix community, then again maybe a year ago on a different community, and even more recently on my self-hosted instance. Don't understand how you can push a chat platform that effectively doesn't deliver ~1/12 messages to random users and let that issue hang around for years.

XMPP looks really interesting as an alternative. Hope that development continued at a brisk pace.

[–] RadDevon@lemmy.zip 6 points 4 months ago (1 children)

Not many... but this community isn't for those people. It's for people who are already predisposed to self-hosting software.

[–] RadDevon@lemmy.zip 36 points 4 months ago (11 children)

Couldn't the same be said for just about any self-hosted app? You can watch video files with a local video player, so no need for Jellyfin; you can save passwords in KeePass, so no need for Vaultwarden; etc.

Seems to me like, if you'd like to have access to this app along with your data from any computer without having to overlay a separate data syncing solution and install a local app on each of those computers, that's justification enough. Or maybe I'm just not understanding your critique here...

[–] RadDevon@lemmy.zip 3 points 4 months ago

Ah, true! Unfortunately for the anonymous LLM whose reputation is at stake, this was something like a platformer.

[–] RadDevon@lemmy.zip 3 points 4 months ago (2 children)

I found a review summary that said the background music in the game made it difficult to see enemies and that I should turn down the BGM track to fix it. 😆

[–] RadDevon@lemmy.zip 9 points 4 months ago

Another frequent transit user here. When people complain that I'm early for something, I like to tell them that, since I ride transit, my choices are to be early or late, but I can't choose to be on time. 😅

[–] RadDevon@lemmy.zip 0 points 6 months ago (1 children)

I’m running this. There are a few things I don’t like about it. The biggest issue for me is that it touts itself as a privacy-centric file converter, but then it makes requests to Google Fonts and Cloudflare. I’ve blocked loading of those scripts in my browser, but I don’t understand why you would add those things to a service that’s supposed to be focused on privacy.

The other issue I had is that the default video conversion server URL is baked into the Docker image. Whereas normally I would configure something like this by simply passing an env var through to the container, here I have to build my own image which makes updating the container more of a hassle.

Seems to be fine as a file converter though.

[–] RadDevon@lemmy.zip 4 points 7 months ago (5 children)

This may be a controversial inclusion, and it’s based on my relatively unsophisticated understanding of Linux. I believe the reason casual computer users hate Linux (generalizing here) is that “Linux” is not one thing.

Commercial operating systems are monoliths. Windows 11 is Windows 11. macOS is macOS. Apart from a few surface-level settings, all instances of them are the same. If you know how to use that operating system, you can go to almost any computer running that OS and start using it, just like you use the one you have at home.

“Linux” is entirely modular. There’s no single thing called “Linux.” You can pick and choose each component to build up your own customized OS from the ground up, and distros take advantage of this. I know just within my household, I have three Linux systems, and casual usage varies wildly across the three. One is a SteamDeck, which is a different kind of thing, but if I just take the two computers as an example, on one, you have an application menu in the top left where the other has an application menu in the bottom left. Also, those menus look completely different. That alone is enough to frustrate a casual user. Now take the fact that they each have different settings panels, different bundled apps, etc. and you have a recipe for making users always feel lost when moving from one system to another.

I don’t think this means you need to teach how to use every available desktop environment, window manager, or sound settings panel, but I do think it would be useful to introduce this concept as part of your curriculum. The sad part is that I think a lot of your audience will tune out at this point because they never had to know that on the commercials OSes, but I think it’s important to be forthcoming about it rather than having your audience blindsided by it.

[–] RadDevon@lemmy.zip 0 points 7 months ago

If I tap-and-hold the image of an image post and tap “Share image,” it shares the image with a URL, which in Signal results in a message containing only the URL. If I use the share button instead of tap-and-hold+share, it shares the actual image. Why are the behaviors different? Why would sharing an image ever lead to anything happening other than just sharing the image?

This is Voyager, btw.

15
submitted 11 months ago* (last edited 11 months ago) by RadDevon@lemmy.zip to c/linux@lemmy.ml
 

I'm on Bazzite Linux 42 and was having some trouble with my 2.4GHz wireless keyboard disconnecting, so I decided to replace it. The new one is having similar issues despite being a different brand (new: XVX, old: Royal Kludge), so I suspect the culprit may actually have been software all along. I have a 2.4GHz wireless mouse connected to the same system that is generally reliable, so I don't believe it's an issue of 2.4GHz interference. The keyboards work well when connected to my Mac, so I don't believe it's faulty hardware.

This keyboard has one feature that may be helpful in troubleshooting: it flashes an LED when it’s trying to reconnect. (The previous one had no indicator.) I can clearly see that, after the keyboard has been idle for a bit, it starts trying to reconnect again. I suspected a power management issue, but I believe I’ve disabled that. I started with a rule in /etc/udev/rules.d/:

ACTION=="add", SUBSYSTEM=="usb", ATTR{idVendor}=="1038", ATTR{idProduct}=="1830", TEST=="power/control", ATTR{power/control}="on"
ACTION=="add", SUBSYSTEM=="usb", ATTR{idVendor}=="0c45", ATTR{idProduct}=="fefe", TEST=="power/control", ATTR{power/control}="on"

(These rules disable power management for both keyboard and mouse, just in case.) I got the IDs with lsusb. I’m assuming the part of the ID before the colon is the vendor ID and the part after is the product ID.

That didn’t seem to help at all, so I tried disabling USB power management with rpm-ostree kargs --append-if-missing="usbcore.autosuspend=-1". That made the problem better, but now it just seems to take longer (a couple of minutes) for the keyboard to lose connectivity. Also, now when it loses connectivity, it seems even disconnecting and reconnecting the dongle doesn't always fix it.

Anyone have ideas what I might try from here?

 

I'm running a Docker-based homelab that I manage primarily via Portainer, and I'm struggling with how to handle container updates. At first, I had all containers pulling latest, but I thought maybe this was a bad idea as I could end up updating a container without intending to. So, I circled back and pinned every container image in my docker-compose files.

Then I started looking into how to handle updates. I've heard of Watchtower, but I noticed the Linuxserver.io images all recommend not running Watchtower and instead using Diun. In looking into it, I learned it will notify you of updates based on the tag you're tracking for the container, meaning it will never do anything for my containers pinned to a specific version. This made me think maybe I've taken the wrong approach.

What is the best practice here? I want to generally try to keep things up to date, but I don't want to accidentally break things. My biggest fear about tracking latest is that I make some other change in a docker-compose and update the stack which pulls latest for all the container in that stack and breaks some of them with unintended updates. Is this a valid concern, and if so, how can I overcome it?

view more: next ›