Hacker News Reader

Sure. Because it's enabling you to establish such tunneling in the 1st place, when working from behind private/non-routable IP-addresses behind NAT.

Without that you wouldn't have any performance at all, in that case. If you're using public IPs you simply don't need it, because no NAT is involved.

View on HN

I don’t get what you mean:

Machine A -> public IP Machine B -> behind NAT

You can always SSH from Machine B to Machine A and expose any local Machine B port to the internet as a port on Machine A. The feature is called GatewayPorts and is a standard SSH feature (at least in OpenSSH). It does not require installing anything other than SSH on Machine A (or even Machine B, unless you want a script to re-establish the tunnel upon disconnect).

Anywhere, Anyplace, Anytime -> public IP running that thing and pointing to -> that thing running behind your dynamic and NATed IP -> golden/Bob's your uncle/gotcha covered/whatever

Reply on news.ycombinator.com