Copied from the original post:

I’m a big fan of automating as much as I can.

• Lights automatically turn on when someone enters the room, but only if it’s dark enough to need lights in that room at that moment. Turn them off automatically when presence is no longer detected (not just IR based motion detection).
• My old washer and dryer alert me when they’re done thanks to vibration sensors.
• Media downloads when added to a watchlist.
• Regular backups locally and to the cloud.
• My phone enables/disables rotation lock based on the app I’m currently using.
• Phone also opens various playlists when connecting to my car stereo based on date/time.
• Various “scenes” to turn devices on and off, control brightness, volume levels, etc. This includes controlling devices that are IR, RF, Bluetooth, and Zigbee based.
• Game servers that update when new versions are released, but only when no players are connected. If anyone is connected, sends a message to a discord server (that is also connected to Matrix) alerting everyone to the available update and asking players to log off at their earliest convenience. Players can also check on game server status with bot commands in the game’s channel on that server.

Everything runs locally and has a manual backup so I can still control everything the old fashioned way if my phone is dead or if my non tech savvy parents are over and need to operate any of it.

Haha no worries, enjoy your coffee!

16 is absolutely not turn based, I’m not sure where you got that idea.

I recently set up something similar to this. I can’t comment on your specific hardware, but I was very frustrated with the limitations of TrueNAS and ended up using Debian and Cockpit with BTRFS for the drives.

I started with two 18TB drives with no RAID, and have since added two 26TB drives with everything’s using RAID1 and ~45TB of usable storage. Converting and adding drives was very simple, but also time consuming of course.

I personally use btrbk with a custom built systemd service and timer. Right now it’s very specific to my infrastructure, but if enough people request it and I have time and opportunity, I’ll post a generic solution here as soon as I can

Look into BTRFS. I’ve been using it for a few months now and it’s awesome. Live disk images with delta changes (saving on consumed space and backup time), even with encrypted drives, and it’s used extensively by Google and Amazon so it will very likely be supported and maintained for a long time to come.

Reading through this thread I’m starting to feel like I’ve probably been overly paranoid.

I have:

• Multiple mobile devices on two different carriers, including computers, phones, and tablets with 5/4G and some with satellite service.
• A high speed wired primary network (2+Gb/300+Mb DOCSIS).
• A robust wireless network (Wi-Fi 7) with multiple access points secured with WPA3 and 802.1X.
• A 24U Server infrastructure for games, networking, media, and Linux ISOs.
• Battle tested 3-2-1+ backup solution, including on site backup with redundancy, as well as two off site backups with redundancy.
• Dual UPS units connected to independent 20A circuits.
• Dual Gasoline and Propane powered backup generators connected to the UPS units to kick in automatically.
• Edit: I forgot to mention that my internal network is running at 10Gb.

Some of that infrastructure was necessary for my line of work during the worst of the covid pandemic, but now it’s more “nice to have in case of apocalypse” equipment.

As others have said, there isn’t much story for most of the game outside of subtle, environmental story that can only be inferred. The biggest story beats happen at the very beginning and very end of the game.

Those callbacks that are annoying rehashes for you are intended to be nostalgic for players of the first game. Keep in mind that there was an 8 year gap between console metroid games.

If you aren’t having fun playing the game, then it’s failing at its primary goal. Feel free to skip the rest of the game, but please be sure to at least watch an online video of the ending, before, during, and after the final boss. The opening of Fusion and other future games in the series reference that ending, and it’s absolutely an iconic piece of gaming history.

Edit: I couldn’t find any examples on PeerTube, but this video of the ending from YouTube is fantastic (should you choose to skip the game).

https://youtube.com/watch?v=rAoVPsjEcNg

One I haven’t seen mentioned, and likely isn’t very famous anymore is (212) 736-5000.

How does this differ from something like FreeMcBoot?

I get that this is better than the old titles.db + specific PS1 disc method, but FreeMcBoot memory cards already made that approach obsolete years ago. Is it just that it supports more models than the older exploits?

Happy to help!

You can find a bit more information at the URL below, and feel free to message me if you run into any issues getting it set up.

https://letsencrypt.org/docs/challenge-types/#dns-01-challenge

Cheers to your journey so far, and to your continued success!

You can absolutely do free SSL certs with Let’s Encrypt without exposing your infrastructure to the internet. Just use DNS based validation instead of HTTP, copy the required TXT records to your domain as instructed, wait for any cache/TTL of any old records to expire (generally 1-2 hours by default), and finally complete the validation.

You’ll need to renew the certs every 3 months, which could be annoying if done manually. If your Registrar has a decent API, writing a script could be a fun automation project. Alternatively I can also send you scripts that I used to use for that purpose.

As a sanity check, I just completed the same setup that you described (Ubuntu Server 24.04 running in a Proxmox VM, Domain name pointing to a CNAME that points to the Dynamic IP, using the installer script, enabled CrowdSec, etc.), and everything worked out of the box. A couple of things I noticed that would also be worth checking now that I’m more familiar with this specific setup are:

• In the config/config.yml file, verify that the dashboard_url is set to the FQDN of the full URL (e.g. pangolin.mydomain.com), and that the base_domain is set to the root/apex domain (e.g. mydomain.com).
• Double check those DNS records. As the haiku goes: it’s not DNS, it couldn’t be DNS, it was DNS. dig pangolin.mydomain.com or dig @1.1.1.1 pangolin.mydomain.com should show the CNAME that points to the A record.
• A 404 page not found error is normal when connecting to the IP address directly rather than using the domain name. I was successfully able to access the dashboard using the FQDN from a local and external network. Depending on your network, you might want to set up a local DNS record that points to the internal IP address as well (though this should be optional in most cases to my knowledge).

I hope that helps!

The script should take care of that config, but it’s something to check just in case there was a typo or anything else like that.

Did you check to make sure the DNS records are resolving properly?

According to the docs, you should be using UDP port 51820 (unless you changed the port in the docker compose file).

You should also check the dynamic config file to be sure that it’s using the correct domain name. See this page: https://docs.fossorial.io/Getting%20Started/Manual%20Install%20Guides/docker-compose

If you’re still having issues, make sure the containers are running with docker compose stats and check the logs with docker compose logs -f. It might also be worth checking the domain name to be sure that it’s resolving to the correct IP address, both locally and externally.

No worries! Yes I’m absolutely still interested, and I have the optiplex set aside for you.

I currently run it with Keycloak for Auth and previously had it behind a Nginx Proxy Manager reverse proxy, but have since switched to using a Cloudflare tunnel.

It works great and allows me to provision limited and controlled access over various game servers to admins of those servers. They can access what they need and nothing more, and only on the servers that they have been granted access.

I use Apache Guacamole, which works great for just about any kind of remote access and has a dead simple to use Docker Container.

It supports folders, copy/paste, uploading/downloading files, multiple open connections at once, and alternative mouse modes for touch screens. Best of all, it’s completely free and open source.

Imagine if OP asked: “I want to repair/maintain my own car, but I don’t want to open the hood or get under the car. What are my options?”

Obviously there would be some options, but those options would be very limited and not ideal. This is very similar. Self-hosting, like self-repair of a vehicle, requires some foundational knowledge and understanding of your specific hardware, usecase, and needs, as well as the knowledge and ability to bring those things to fruition. There is no single universal answer that applies to everyone, but those skills can be acquired by anyone.

I don’t think self-hosting is any more doomed than self-repair of a vehicle. It’s certainly not for everyone, but it can be satisfying, rewarding, enjoyable, and generally optimized for those who choose to participate.

Interesting and good to know, thank you for the info! I haven’t run into that issue with synapse yet, but I’m also running it in a docker container on an enterprise server. I also don’t currently have any chat rooms with more than 5 or so users, and I imagine that is also a factor.