lckdscl [they/them]

I’ve been using Authelia with several OIDC integrations for a while now. Works great. They’ve released a huge update like a day ago too. Out of the ones you listed, it’s very lightweight too. The docs are a bit all over the place but it is quite comprehensive.

I did look at Zitadel and tried setting it up myself but I just couldn’t get it to work. The docs are a bit vague.

If you just want to view logs, then a lightweight viewer I really recommend is Dozzle.

I’ve tried nearly every selfhosted dashboard out there and in the end settled for static html/css/js. If you want to access links quickly by typing abbreviations then use something like https://github.com/Ozencb/tilde-enhanced. A lot lighter and can be used with an existing webserver too.

https://tailscale.com/kb/1054/dns#nameservers

and

https://tailscale.com/kb/1114/pi-hole#step-3-set-your-raspberry-pi-as-your-dns-server

Set tailscale to use your dns server to resolve your services (or all traffic if you prefer). Assuming your dns server is on 100.x.x.1:53, then put 100.x.x.1 as a nameserver.

How about Uptime Kuma status pages? They’re separate from the admin page and you can add Docker containers as monitors.

SFTPGo supports OIDC and has a lot of ACL features. It allows users to have their own folders, as well as shared volumes between a group.

I like Halls of Torment and Boneraiser Minions, both really nice on the Steam Deck.

This might be an issue with opensearch.xml, which is a standard for how browsers recognise search engines.

See here:

https://github.com/hnhx/librex/blob/main/opensearch.xml.example

I don’t know how you’re hosting it, but when I was hosting LibreX, I had to make an opensearch.xml with the correct domain and bind mount it to the correct location. I don’t exactly remember the details since I moved to Searxng.

Also, if you’re not aware, LibreX was forked to LibreY, which is the updated repo.

Navidrome replaced Spotify for me, with Symfonium on Android, I’m never going back. On PC you can use any Subsonic client, and there are plenty I threw Tailscale on top to access it when I go out.

Org-mode, with Orgzly on Android, sync via a WebDav server, which you can also mount on you PC and literally use any editor to edit.

Well guess I’ll add one more to my three a day.

Okay I think I might know what you mean? I just tried doing that and got it to work. We can compare what we did. Here’s mine.

I created a shared folder called “Shared”

then I create a group called “All” and mount the “Shared” folder to /shared

I went to a user and add them to group “All”

Examining that user’s files

I can navigate into that shared folder and access everything (I have stuff in there already).

To set up the folder, which I called “shared”, I set the home directory for it to /srv/sftpgo/data/shared. For reference, my user home directory is /srv/sftpgo/data/user1. Then to allow user1 to access it, I mount it as a virtual folder. Is this what you did?

I know you meant well, but I don’t think their interpretation implied any logical fallacy. I used a conditional statement but my statement was prescriptive, not descriptive.

The difference between “I should” and “I have to/must” is a modal one. I implied “if I have to X then I shouldn’t Y”. They swapped X and Y around to get “If I have to Y then I shouldn’t X”, which is just a plain misinterpretation. The use of what is and what ought implies a recommendation or opinion, not mutual exclusivity. For that, I would have to use the same modality “If I have to X then I must not do Y”.

It’s like mixing up “If I have an infectious disease, I shouldn’t go outside” vs. “If I have to go outside, I shouldn’t have an infectious disease”. To me, they have a subtle difference. There is compromise and decision-making involved.

I’ll spell it out anyway because why not. I can’t be bothered to edit my original comment. While it’s sensational-sounding, anyone who take issue with what I said don’t take surveillance properly so I can’t help them, while those that misinterpreted me like nous did can find out for themselves here.

Yeah I agree. To be clear, if you take the reverse of my statement, i.e. if you’re on Windows, you shouldn’t use Tor, then I would be gatekeeping.

But I’m not implying that, but rather the reverse. I’m saying if you have use Tor for whatever reasons to bypass censorship, do illegal stuff and avoid being tracked, you should at least be aware that at the kernel level, how you’re accessing the internet has already been compromised by Microsoft, and consider alternatives OSes

Of course I’d still want people running Windows to be able to use Tor, and also I’d say leaving Windows isn’t something you would only do at the “highest threat model”.

Privacy will almost always be a trade-off with convenience, I’m pushing the awareness to get people to act, should they choose to. That’s all.

Agreed. I thought of ISP restrictions too, but I would say if where you live places a level of censorship due to political reasons or otherwise and you need to access it for whatever reasons so you need Tor then by all means Microsoft is not your friend since they’re a privacy nightmare.

There are also VPNs for banned media, I typically wouldn’t want to use Tor for anything more than textual content as it puts too much load on the Tor network.

If you have to use Tor you shouldn’t be using Windows.

+1 for Netdata, very fast and a lot of alerts have already been set-up. It also has a lot of plugins, as well as the ability to use Prometheus metric endpoints. The local dashboard is near parity with the cloud one, and setting it up is as easy as running their bootstrap scripts. There is decent documentation too, if one gets stuck.

Subsonic-based alternatives are good too. Navidrome and gonic, for instance.

I think all the RAM related issues were closed a while back and were supposedly fixed. I just don’t understand why when interfacing with the front-end, it uses so much it would get OOM kill itself with 1.5 GB allocated memory.

Every page, as well as loading in the initial dashboard from an idle state, spikes the RAM. Are there no clever lazyloading happening or something? Surely viewing and modifying database entries can’t be this memory intensive?

Maybe it’s just an unoptimized Python thing. I stopped self-hosting stuff written in Python, with the exception of Linkding (which takes a while to also submit a link) and Whoogle.