My homelab has been mostly on autopilot for a while. Synology 6 bay running most lighter weight docker stuff (arrstack, immich, etc) and an Intel nuc running heavy stuff (quicksync transcodes for Plex+jf, ollama). Both connected to digitalocean via WG for reverse proxy due to CGNAT.

I had my router SSD either die or get corrupted this past week, haven’t looked much at the old SSD besides trying to extract the config off of it. I ended up just fresh installing opnsense because I didnt have any recent backups (my Synology and nuc back up to rsync.net, but I haven’t gotten around to automated backups for my router since it’s basically a plain config, and my cloud reverse proxy which is just a basic docker compose + small haproxy config). Luckily my homelab reaching out to the cloud reverse proxy means there’s basically no important config on my router anymore, they just need DHCP and a connection.

Besides that the arrstack just chugs along on its own.

I recently figured out I can load jellyfin playback URLs into vrchat video players, either direct stream or through the transcoding pipeline as an m3u8 that live transcodes based on the url parameters you set. This is great because the way watch parties in VRChat works is that everyone in an instance loads the same URL pasted into media players and syncs the playback. That means you need to have a publicly accessible url (preferably with a token of some sort) that can be loaded by an arbitrary number of unique IP addresses simultaneously, which I don’t think is doable with Plex.

I’m now working on a little web app to let me log into Jellyfin, search/browse media, and generate the links with arbitrary or pre-set transcode settings for easy copy/pasting into VRChat. The reason it’s needed is that Jellyfin only provides the original file without transcoding when you use the “copy stream” option, so I believe the only way to get a transcoded stream url currently is to set the web interface to specific settings and grab the URL from the network. But that doesn’t let you set arbitrary stuff like codecs and subtitle burn in and overriding what it thinks you support. So a simple app to construct the URL will make VRChat watch parties a lot easier.

The new beta timeline is sooo smooth! I finally don’t hate scrolling back to find a specific old photo. The scrolling performance feels completely native to me now.

I think what you want is an EDID emulator with passthrough or whatever it’s called. EDID is how a monitor tells a device what resolution to send and other info. Some cheap HDMI splitters, adapters, audio extractors, etc will let you emulate a specific EDID. One of my audio extractors lets you fake stereo vs surround support to trick the source into sending surround - I think that’s also through EDID - since if you’re trying to extract surround, it might be because your real TVs EDID is for stereo I assume. So you probably want something like that in before the switch so that the laptop always thinks something is plugged in. Your switch seems to be too smart in actually passing through the real monitor’s EDID so the laptop is able to see when it switches.

Open webUI connected to ollama can do this. In openwebui, if you edit any one of your responses, it forks the conversation. You can flip between each branch using the arrows below any of your responses. If you click the 3 dot menu and click overview, it opens a graph view that shows the branches of the conversation visually.

As a point of reference, I have a 5070 ti oc (300W tdp, suggested PSU 700W according to techpowerup) with a ryzen 7 7700 (65W tdp) and I use a Silverstone SFX 700 W 80+ platinum and it works great. I’ve monitored the GPU wattage and it generally doesn’t go above 200ish in practical usage.

Fwiw Anubis is adding a nojs meta refresh challenge that if it doesn’t have issues will soon be the new default challenge

Archive link for the Bloomberg article in case the gift link stops working https://archive.is/2mltm

deleted by creator

Imo that’s perfectly fine and not idiotic if you have a static IP, no ISP blocked ports / don’t care about using alt ports, and don’t mind people who find your domain knowing your IP.

I did basically that when I had a fiber line but then I added a local haproxy in front to handle additional subdomains. I feel like people gravitate towards recommending that because it works regardless of the answers to the other questions, even their security tolerance if recommending access only over VPN.

I have CGNAT now so reverse proxy in the cloud is my only option, but at least I’m free to reconfigure my LAN or uproot everything and plant it on any other LAN and it’ll all be fine.

This is 99% my setup, just with a traefik container attached to my wifeguard container.

Can recommend especially because I can move apartments any time, not care about CGNAT (my current situation which I predicted would be the case), and easily switch to any backup by sticking my boxes on any network with DHCP that can reach the Internet (like a 4G hotspot or a nanobeam pointed at a public wifi down the road) in a pinch without reconfiguring anything.

It’s time to get with the RCS program, already!

Tell that to my MVNO 😮‍💨

It completely depends on the project and the maintainer. Lots of repos are just source files, some include instructions on how to compile, some have no instructions. Some post releases with a pre compiled executable that you can download and run, others post releases that just have a zip of the source code. Some projects use GitHub pages to host detailed manuals, tutorials, etc.

If you share a link to a specific project, people may be able to help you get it running. Whether or not you need a bunch of tools like an IDE just depends on the project.

Already on my wishlist apparently, I remember seeing some early gameplay footage and being sold on multiplayer combat with a ship that seems to act like a mobile base.

I’ve been pretty happy with rsync.net, with the promo rate I’m on I get 1.6TB for $10/mo. More than enough for backing up my most important files, private keys, camera roll, anything irreplaceable.

The rest of my NAS is 50TB of movies and shows so I’ve just resigned myself to the fact that it will never be affordable to backup more than my user folder.

!?!?!?!?!?!?!?!?!?!?!?!?!?!?!?!?!?!?!?!?!?!?!?!?!?!?!?!?!?!?!?!?

What model size did you run on your laptop? I have an Intel Nuc with an i7 and I run various models on CPU (it doesn’t have a dedicated GPU) and while I can’t run stuff larger than ~14b or so, models up to around ~7b aren’t too slow. If I try to run a 32b then I get a similar experience to you. I tend not to go below 4b because that’s when it starts being dumb and not following instructions well, so just depends on how complex your task is.

Immich is pretty good for this if you take pictures at each location. It has a global map that shows all your photos with a heatmap-style display and a drawer that shows a grid of the photos within your viewport as you can and zoom around. It doesn’t seem like you can view a specific album on the map currently but you can at least filter the map to favorites or a date range.

I think you can use Immich external libraries for this, also to be extra safe you can just mount your external images folder as read only by adding :ro to the docker volume mount so that the container won’t be able to modify anything as a precaution.

I use a .dev and it just works with letsencrypt. I don’t do anything special with wildcards, I just let traefik request a cert for every subdomain I use and it works. I use the tls challenge which works on port 443, so I don’t think HSTS or port 80 matters, but I still forwarded port 80 it so I can serve an http->https redirect since stuff like curl and probably other tools might not know about HSTS.

Gotcha thanks for the info! It looks like I would be fine with ocis or opencloud, but since my main use case and pain points are with document editing which is collabora, it probably wouldn’t change much besides simplifying the docker setup (I had to make a gross pile of nginx config stuff pieced together from many forum help posts to get the nextcloud fpm container to work smoothly). But it already works so unless it breaks there’s little incentive for me to change.