BreakDecks

No shit the VPN requires an open port, I never said otherwise, but if your router is the one running the server, you aren’t forwarding the port. The router itself is listening on its WAN interface.

The VPN prevents you from having to forward any ports, because the router allows you to tunnel in. The only open port will be whatever port the VPN server listens on, and it isn’t a forwarded port.

Source: I literally work at a VPN company.

going from a public device to a device in a protected network

You mean the literal function of a VPN?

A VPN eliminates the need for port forwarding.

Edit: Not talking about a “privacy VPN”, but an actual VPN that lives up to the name “Virtual Private Network”, where you are connecting to the private network you wish to access.

There’s an AliExpress link in the article that clearly prices it at $260…

Not if HIP is a part of your network stack (adding a waist). Unfortunately this is not a plug-and-play solution unless you have the budget for my employer’s product, which is not priced for consumers.

https://lemmy.ml/comment/9601780

Oh man, uh, here’s a primer: https://mkomu.kapsi.fi/hipl/index.php?index=how

I am most familar with HIP, but there are ways to isolate hosts so that they can only talk to what you want them to talk to in a distinctly different way than a firewall. You could have three hosts (A, B, & C) on the same subnet where A can talk to B & C, but B & C cannot talk to each other. Likewise, A and C could have access to an Internet gateway, while B does not.

So far HIP is the only protocol I have seen for microsegmentation that actually works in an intuitive way, but I suspect Wireguard could be used to the same effect with some creative engineering.

This one of those questions I am overwhelmingly eqipped to answer, but only with the weird proprietary knowledge about software defined networking and microsegmentation that my job has endowed me with…

So I’ll resist the urge to give you that overcomplicated answer and just say get a firewall like others have suggested.

My favorite part about using an old laptop as a 24/7/365 plugged-in server is the anticipation of when the lithium battery will explode from overcharging.

I don’t know of any existing project, but this sounds pretty trivial to make with Python using http.server and PIL modules.

Upload file and store in RAM. Delete after first fetch. Optional watermarking for image files using PIL.

If you can RDP, just copy and paste the file from one computer to the other.

Never said otherwise, I just want to make sure we’re not scaring people away from Lemmy administration and moderation, as if they were risking going to prison as a child sex offender or something.

Any of filter or image processing technique that fools machine vision.

Example: https://sandlab.cs.uchicago.edu/fawkes/

At a high level, Fawkes “poisons” models that try to learn what you look like, by putting hidden changes into your photos, and using themn as Trojan horses to deliver that poison to any facial recognition models of you.

This could be done with any kind of image or detail, not just faces.

I hate how everything is a double edged sword, because this is now also the perfect tool for making sure your CSAM doesn’t trip the filter. Also, it uses CLIP so a simple obfuscation overlay would render it useless.

the admin will be liable for it.

…

These bad actors can then share these links around and you would never know! They can report it to the FBI and if you haven’t taken it down (because you did not know) for a certain period, say goodbye to your instance and see you in court.

In most jurisdictions this is not now it would work. Even a less tech savvy investigator would figure out that it was an online community not obviously affiliated with CSAM, and focus on alerting you and getting the content removed.

There’s this misunderstanding that CSAM is some sort of instant go-to-prison situation, but it really does depend on context. It’s generally not so easy to just plant illegal files and tip off the FBI, because the FBI is strategic enough not to be weaponized like that. Keep an eye on your abuse and admin email inboxes, and take action as soon as you see something, and nobody is going to shut you down or drag you to court.