Awesome, good to hear. Just a reminder that I forgot when I sent that link: double-check the Tutorial, it might have slightly outdated information. But a few lessons I learned when setting it up:

Pangolin uses its own auth. You can use a OIDC provider if you like, but you have to set it up. I use TinyAuth in conjunction because I have so few users.

To bypass auth on Pangolin, you must disable it and it will appear “Unprotected”. This helps with mobile apps, which don’t know how to deal with third-party login screens, but it leaves the login screen of whatever server you’re running (like Jellyfin) accessible by everyone. So something more hardened as your OIDC provider at the frontend help. Although lots of apps, again like Jellyfin, don’t have native OIDC login.

The solution would be to route it with Private Resources, which are zero-trust connections. Great for a few people on your network, but it requires you setting up the Client app to access the Pangolin network. I’ve had issues with DNS on Android before, but your problems may vary.

If you have any more questions, let me know. I’ve been using Pangolin for only about a year, but I think I learned enough to help.

Sure, https://pimylifeup.com/pangolin-linux/

But I have to say, the Pangolin Docs are quite good as well: https://docs.pangolin.net/self-host/quick-install

I use Pangolin on a kinda cheap VPS on Hetzner (they’re now a bit too pricey), and route my services through Site Connectors (Newt, it’s called). Dealing with opening my firewall is a pain I don’t wanna think about, but Pangolin creates a websocket and can do NAT traversal from the Site Connector, so no swiss-cheesing your firewall.

It’s nice because you can contain different services with VLANs to further segment and protect your network, and each Site Connector can work for many services (if there’s no firewall rules restricting it/VLANs) or you can put a Site Connector on each server individually, if segmentation is important to you.

For something that uses immeasurable amounts of stolen data to be trained on and function, without credit or compensation, and breaks so many laws I can’t list them, I think it’s pretty unethical to use. Or something that burns immeasurable amounts of potable water and electricity to function at all, I think its pretty unethical. For something that artificially generates scarcity for chips we desperately need, I think it’s unethical. For something that authoritarian governments are lapping up as a form of propaganda, disinformation, misinformation, surveillance, I think its pretty damn unethical. All so people can write an email 10 seconds faster, in a tone that’s not yours, to be sent to a AI to answer that email. For the people who say “I think it can be used for good” or “it’s just a tool!” What kind of tool sacrifices so much of which we take for granted? Are these things not important to you? Look at that list again! Fuck, look around, your silly chatbot is not changing the world in the way you want. Maybe you get some minor benefit from your chatbot, but the world shifts further to a dystopian shithole for your marginal gains.

Say you’re in the desert and miraculously you found a glass of water, but it is filled with shit particles. Ya, it’s a glass of water, but it’s filled with shit. And you’re out here thinking to yourself “hmm, there’s water here, and I can drink it. So what if it’s filled with shit? I can use it.” #fuckai

I have successfully set up Synapse Matrix server twice before and it is still a ridiculous convoluted mess. Something like Conduit for Matrix might be better, but progress there is glacial.

The naming conventions nowadays for tech is just the stupidest sounding bullshit. Get the all new AMD AI Mega MAX+ AI Pro Ultra AI Quantum Pro Max.

If, somehow this next midterm election goes well, then maybe there’s a mild chance at a more peaceful methods. I doubt it. It’ll be business as usual in a few years. The next round of fascism will be elected in the next election, another neoliberal supporting genocide will be on the Democratic ticket and another round of madness of fascism ensues. Frankly, the US as a democracy has been officially dead since the Supreme Court gave the president immunity from anything. How can you run a democracy with such power in one person?

There’s undercover thugs kidnapping and murdering people. There’s a president making $6 million a day from illegal trading. There’s a new focus on domestic left wing “terrorism”. But not a peep from the masses. It’s business as usual, as always.

You’re almost there. Fight for democracy. The Nazis are in charge now and they’re fundamentally dismantling our democratic institutions. These CEOs are instrumental in this dismantling, there’s no more peaceful protests that will reverse this course. We’ve fought Nazis before, we’ll do it again.

For people who want an opinionated browser… Yes. They have sponsored shortcuts which cannot be disabled and they ghost out the option to in the settings. If you want to dig around the about:config and tweak things, fine, but I’d rather use a browser I can make my own. Librewolf is excellently bare bones.

Both. There’s a difference between showing some clerk your ID compared to uploading it to the internet. It’s not a question of if it being hacked. It will be. Denial of this is dangerous. If you don’t see this as important, you’re desensitized by the sheer number of yearly cyber attacks.

And that’s only the start. Children will only be marginalized. Protected groups will be increasingly threatened. Take your pick on whatever organization you want to look at, and they’ll say this doesn’t help anyone, except maybe foreign adversaries and hacking groups. What happens when the next government comes along and decides to make a more US kind of implementation? The point is, that we should not make this the precedent. Ever. Kick it while it’s down.

Explain basically every privacy, cyber security and child safety organization saying this is a bad idea, then.