Re: That old chestnut
Only working behind NAT is largely a solved problem, as both TCP and UDP traffic can breach NAT from the inside using off the shelf libraries. This could have been added as a base extension to the IP4 IP protocol 10 years ago, and would have if those that drunk the IPv6 "this is the one true way" coolaid would stop BLOCKING it.
Most of the rest of the argument presented involves the authors apparently delusional idea that we are all going to allow unchecked inbound traffic on arbitrary protocols into our networks. That's not an IPv4 or routing issue. It's a firewalling issue. Moving to IPv6 will not ever "fix" that. I could allow that now on IPv4 if I was either dumb or insane.
We have seen over an over again that there is always an idiot whinging on the internet because someone is blocking their obviously bad idea instead of listening and understanding why they are. The decades of UPNP vulnerabilities we are still mopping up are a great example of where taking this kind of advice gets you.
People will move off IPv4 when they finish fixing it's erstwhile replacement so that is solves issues they actually have, and doesn't cause undue problems they can't work around. The people pushing IPv6 made it to solve their problems, not mine. So they increased the address space to a ridiculous amount, and re-tuned the routing logic to take pressure off of routers. That's great for *NIC registries and backbone operators that literally just push traffic.
Want to get people on IPv6? All it will take is to fix a couple problems and add a little sweetener. UP the minimum MTU on the backbone to support some reasonable approximation of jumbo frames to take pressure off of big TCP connections, and fix the multiple internet connection/failover problem without handwaving some BGP noise that can't even fail over in the order of seconds, let alone transparently.
Until then, the IPv6 traffic I see will continue to be mostly garbage that I should and am blocking, and the IPv6 link will be unneeded pain that necessitates complexity, instability, and unreliability and offers nothing in return. Based on the track record of failure, I expect this to go like the endless wars over CAT 7 or later cable specifications, so I'm not holding be breath.
Biting the hand that feeds IT
