[reincarnate0x14]

UPNP and a dozen other NAT defeating tactics exist and have since the early 2000s. NAT translates addresses. Thinking a non-routable range is safe because it's behind NAT is at this point grossly ignorant of how modern network equipment works. It's kind of like port-knocking; yes it makes the attack slightly harder, but doesn't prevent it.

e.g. symmetric NAT exists and often doesn't come with a stateful firewall. Just because the linux box with iptables is protecting your network uses NAT doesn't mean NAT is doing the heavy lifting here. I can see the OMG MY PRIVACY crew is out in force here apparently misunderstanding that NAT does not do that either. I mean, we can explain things to you, but we can't understand it for you.

[stackghost]

>UPNP and a dozen other NAT defeating tactics exist and have since the early 2000s.

I know that, and you know that, but squillions of people think that turning the UPnP setting off (if they even know what that is) is sufficient, which is why the myth persists.

[fc417fc802]

UPnP is only relevant for software that's already running on your machine locally. People here aren't generally talking about outbound connections and I think you both know that. The practical effect of NAT as commonly encountered in a residential setting is to drop inbound unsolicited connection attempts.

And yes, everyone is aware that you could also do that with a stateful firewall. And no, none of us care about arguments of definition that attempt to frame NAT as technically being a firewall based on how it operates in practice. Being intentionally obtuse by refusing to acknowledge the obvious isn't going to convince anyone.