That would be interesting. If this is going where I think it is then it would probably suck seeing as touch screen input feels exponentially worse with latency but nonetheless would be really cool to see work for hosting a web app of sorts for desktop use.

Check out linuxserver.io, kasm and whatever the x11 version of waydroid is called. I can’t remember it at the moment. That should give you enough to get started building a container if you want.

This is dope! I wish there was a proper community that did tests like this in mass using open source standardized methods/hardware.

Totally agree. It sounds like something was lost in translation here by the final edit of potentially some run though a llm for proof reading to dumb it down enough to either just make it more consumable, more clickbait or realistic both.

My guess is the actual research reported that it was 100s of packets per second (not screenshots) which is still a lot more than you would expect even for spyware. Either way it’s been well known that smart tvs are spyware ridden, I don’t need a paywalled service to tell me that.

Aaaaand

^Pop goes the AI bubble.

Last stages of capitalism for tech is usually in the form of an ipo of some sort which is what this will lead to.

There will be other cool shit obviously with integrations and tools that will hopefully trickle down to open source models but the writing is on the wall. This is a cash out and enshittify move.

The best news out of it is we will start to see less and less “our company is Ai and we shoved Ai into said thing” as the companies late to the game will continue to shoot their shot until OpenAI has completely dominated the market and investors stop caring.

Honestly this sounds like a bit of a pickle. If I were in your situation I would just use one of the cellular carriers 5g internets. I personally use a T-Mobile 5g internet hotspot with a fresh tomato flashed nether 6700 plugged into it. Then I basically do all of my networking from that. Latency is a fair bit higher (usually about 30-50ms) but upload is significantly better than spectrum.

Best option I have found so far.

You make good points here for the beginner however there are better alternatives and solutions for basically everything you mentioned here. The biggest I want to address is conflicts on your system. Generally running servers on metal is just outright bad practice. Containerize. Always containerize. There are lots of great options. Docker, podman, Lxc, helm, flatpak… hell. Snap if you must. Running servers on metal is generally is just asking for trouble unless the system’s entire purpose is for that. Also the cg-nat situation. Personally been behind it for a few years but it’s not a problem as long as you have a reverse proxy tunnel in place. Not a hard fix at all.

For those who are saying just photoshop.

https://methodshop.com/photoshop-money/

Obviously you can just use other software but PS is the main choice for image editing. What they need to do is put legislation in place and it will make the biggest players implement this form of drm.

They are deafening but usually they are very well insulated seeing as keeping servers cool is very expensive and extremely important.

Jesus Christ. Have people never heard of covers? Every song here is in some way or another akin to a published cover of another song. Pretty bad ones at that. Obviously if it were matching the songs one for one, then it would be considered copywrite enforceable but realistically these would be more along the lines of copywrite abuse. The music labels would absolutely love for this precedent to be set so that anything even that remotely resembles anything ever made will allow them to own new independent artists within established genres.

The cases

Here is a list of cases that set precedent. The thing that connects them all and makes them relevant is that the defendant was either successful, made a lot of money, was very popular or it was the label attacking a artist for sounding like themself after leaving the band. See John Fogerty v. John Fogerty

But duck duck go has ads… They are just as search results. My Pihole will actively block access to those links because they are promoted content. They also have a little “ad” tag alongside the post.

Tons of good responses here. I’m surprised that nobody has brought up Tailscale though. It’s def the easiest vpn solution I have found. It’s got some great documentation and how to projects to get a home lab running and it’s got its own domain system baked in most of it being zero configuration. You can access mullvad vpn exit nodes straight from it, and set up those domains with ssl super easy e.g.

sudo tailscale serve —https=443 localhost:8096

That single command would allow any other devices connected to your Tailscale account to reach your Jellyfin using the domain “{serverhostname}.[tail-scale].ts.net” complete with a private reverse proxy and ssl cert.

There are a few things to click around in tailscale on but it’s a extremely easy to use free application that has made my self hosted life significantly easier due to my system living behind multiple firewalls that I sadly have no control over.

According to ordinary things it takes hours.

https://vid.puffyan.us/watch?v=eeurjIY3__o

https://www.youtube.com/watch?v=eeurjIY3__o

I love Jami, that being said it has one massive problem. In order for it to be usable on local networks you need to either port forward the peer to peer port, set up a proxy relay or use the proxy relay that Jami provides. That’s not a big deal to set up or make any of those changes but they are things that need to be done. There is no real warning about it and when you are using mobile it works just fine due to cg-nat so the problem ends up seeming intermittent. Like I said I love Jami but I don’t think it will ever really be a contender for a mainstream chat platform unless they make some pretty big changes to how relays are handled or become more transparent about this particular problem in the setup process.

That being said… Matrix is pretty rad. Like really really rad. Go look at that. It feels a lot more like a federated chat service because it is designed from the ground up to be that. Plus interoperability with clients is cool. Plus if you set up your own server then you can add bridges to sync all of your accounts to use matrix so that you don’t have to force anyone to leave their respective platforms and you can have one unified repository for all of your messaging. Basically means you get to use what you want and other people can use what they want. Go look at it now. Go on git.

https://matrix.org/

I think something that a lot of the comments are missing here is the fact that threads, Instagram and Facebook all have been migrated from individual accounts to ‘meta’ accounts. I’m certain that we will see this happen with many platforms unless there is a serious shift in data protection laws. I don’t personally think it’s great that it’s the case but that’s just how it is. The meta platform is quite similar to how google migrated YouTube users to google accounts way back in the day. This monolithic structure ensures that they can keep your user data in a more streamlined database. From a sys admin and a business perspective it makes a lot of sense. From a user who doesn’t care and already uses all of those services perspective it makes a lot of sense. From a privacy conscious user perspective it makes no sense. Then again metas platform is in no way for the user who cares how their data is being handled.

I guess another perspective is talking about interoperability. It kind of feels like they are taking the web3 (I know it’s a loaded term) approach but instead of applying it in a way that allows free development and communication in a way that basically pulls from decentralized/distributed databases you instead get a centralized monolithic model that creates interoperability within their own walled garden.