Hacker News Reader

Hacker News

Comment by an_d_rew | Hacker News Reader

Parent Full thread

an_d_rew

Thank you!

I know it’s all the rage to say “gee why no v6 yet?”, but that’s a LOT of infrastructure and testing to overhaul…

The effort is much appreciated!

View on HN

I don’t really understand the benefits, would you mind explaining them to me ?

> I don’t really understand the benefits, would you mind explaining them to me ?

Want to run a public service with end-to-end connectivity? Go to your RIR to request an IPv4 block and be put on a waiting list.

Or break out your cheque book and be prepared to cough up $35+/IP for the privilege:

* https://auctions.ipv4.global

* https://ipv4marketgroup.com/ipv4-pricing/

* https://ipv4connect.com/marketplace

Or get an IPv6 address block right away for $0.

You'd still need IPv4 to connect with the Internet.

With the IPv4 internet, that is. You don't need IPv4 to connect to most GitHub pages anymore.

Admittedly, IPv4 internet is a good chunk of the Internet, but maybe not the part of interest to you at a given point in time. http://www.delong.com/ipv6_alexa500.html

Plenty of folks are IPv6-only. Mobile on T-Mobile for example:

* https://www.youtube.com/watch?v=nNMNglk_CvE

the Internet is almost out of IPv4 addresses, and the ones that are left are becoming expensive to obtain. Rather than hide whole blocks of users behind NAT, they can just use IPv6.

>Rather than hide whole blocks of users behind NAT,

... which creates all sorts of routing issues.

If you don't have your own publicly accessible IP address it creates all sorts of connection issues.

I don't want to have my own IP, at least not without frequent changes. My servers should have static IPs but having static IPs at home or for mobile devices only makes tracking easier..

for some use cases. for example, most home users are behind ipv4 NAT because of their wifi routers. nobody really notices a problem.

Mostly because we've invested huge amounts of engineering effort to work around the problems. The result is that for the most common use cases, things mostly work.

NAT systems are optimized for a few devices to be active at a time. As the number grows they might not co-operate well. Game consoles are infamous for networking problems when you have anything other than a single machine.

And the Wifi router is at least under my control (although depending on where you live and which ISP you use even that isn't necessarily true), so I can still set up some port forwarding for my internal devices as required. Because of the shortage of IPv4 addresses however carrier-grade NAT is becoming more and more common, which is truly an abomination.

The big one is the address space expansion, which in addition to solving "can't get addresses" also makes routing cheaper and faster (can allocate addresses based on network topology, rather than on what's available). Whole they were breaking backwards compatibility anyway, they also introduced a faster way of getting initial addresses, changed headers and dropped some features to make routing easier to implement in hardware, and added some features to make annotation of packets in the network backbone during transit easier.

IPv6-only hosts will be able to connect to github pages now, if configured properly.

There are a couple use-cases for that, but it should allow cheaper VPSes, or less complicated network configuration by eliminating NAT.

If a host only needs to access GitGub and a few select hosts, it can be granted an IPv6 without setting up DHCP, NAT, IPv4 routing tables, etc.

I've been considering using NAT64 as a way to eliminate the need for dedicated IPv4 configuration at home, unfortunately I recently moved, and the current provider is IPv4-only. This infuriates me to no end.

Reply on news.ycombinator.com