The Future Without IPv6

May 23, 2008 11:59 PM

The Future Without IPv6"...imagine you're in the business of squatting on domain names today. It's pretty easy to see that a market is going to be opening up soon allowing you to speculate on the future value of IPv4 address allocations. What would you do? You'd be trying to eat up as much of that free pool as you can before it's all gone."
posted by gsb (29 comments total)

Here's a comprehensive wikipedia article about IPv6 that explains the background a little bit.
posted by gsb at 12:01 AM on May 24, 2008 [1 favorite]

Can someone explain the author's last paragraph to me? Specifically:

Now, IPv6 proponents think that the cost of maintaining IPv4 address allocations will drive the transition to IPv6. I'm not so sure. In fact, I'm thinking that's Yet More Kool-Aid. There is no cost for IPv4/NAT high enough to drive adoption of IPv6, because IPv6 will never be an alternative to IPv4. The Internet will turn out to be, always and forever, the IPv4/NAT-only Internet we have today.

This sounds suspiciously like a circular argument to me, but I don't know if there's a specific technical meaning to "IPv6 will never be an alternative to IPv4" that I'm missing.
posted by chrominance at 12:34 AM on May 24, 2008

IPv6 in Japan has been the coming thing since, oh, about 1997, if memory serves me. It's still coming, despite government money being thrown at it there, and a fair amount of private money, in the form of mobile telephone network development.

If they figure out how to use IPv6 to make something, the rest of the world will follow. If they don't, eventually, China or India will. But I doubt it will be due to exhaustion of IPv4 address space considerations. More likely, it will be as a result of some killer app that exploits IPv6 end point addressing, security, or multicast support. IPTV, may, in fact, be the killer app, as a multicast service, that gets IPv6 off the ground, finally.

But heck, I've been saying that since 2002, too.
posted by paulsc at 12:48 AM on May 24, 2008 [3 favorites]

If they figure out how to use IPv6 to make something, the rest of the world will follow.

You mean IPv6 porn?
posted by rokusan at 12:58 AM on May 24, 2008 [2 favorites]

The address space exhaustion problem in IPv4 leads to another, potentially even worse problem: extremely complex routing situations, especially when you have to try to treat non-contiguous addresses as a single block. At some point you'll see people running servers behind nested full-cone NAT gateways (where each server is mapped to a unique IP outside the NAT), just to keep the routing situation simpler (like if none of your IPs were contiguous).
posted by blasdelf at 1:00 AM on May 24, 2008 [1 favorite]

"... just to keep the routing situation simpler (like if none of your IPs were contiguous)."
posted by blasdelf at 4:00 AM on May 24

That was what CIDR was supposed to do, but frankly, supernetting never really caught on, as widely as many expected. Still, I think all the backbone routers and ISP class equipment in the commercial Internet are CIDR ready, should that need come to pass.

What's really developed, instead (at least in North America and Europe), is an Internet that is largely hierarchal from a routing standpoint - a lot more like the old switched circuit telco routing model, than the packet switched model of a network that automatically routes around damage.
posted by paulsc at 1:09 AM on May 24, 2008 [1 favorite]

Of course, you could just turn off IPv4 and see what happens: "Yesterday evening, shortly before 7 PM, the mixed IPv4/IPv6 network was turned off, forcing the IETF to proverbially eat its own dog food and talk to the world outside the meeting hotel through the remaining IPv6-only network." (Spoiler: Embarrassingly, DNS breaks. You shouldn't need DHCP with IPv6, but without it you have to manually configure a nameserver address. And Windows XP is even more broken than that.)

I attribute part of IPv4's inertia to the mystical security properties attributed to NAT by so many networking/IT "professionals."
posted by sdodd at 1:34 AM on May 24, 2008 [1 favorite]

I attribute part of IPv4's inertia to the mystical security properties attributed to NAT by so many networking/IT "professionals."

I kinda like being able to memorize, or at least recognize an IPv4 address, myself. I don't look forward to the days of scanning logs full of "fe80::200:f8ff:fe21:67cf" and wondering who the heck that is.

At least when I see "74.53.68.130" a hundred times per day, I know what it is.
posted by rokusan at 3:05 AM on May 24, 2008 [1 favorite]

I still believe singing a song it's the solution and the cause of any problem in the world.
posted by darkripper at 3:19 AM on May 24, 2008 [6 favorites]

darkripper, that was... TOTALLY. AWESOME.
posted by sdodd at 4:36 AM on May 24, 2008

darkripper: Great link, although I died a little when why spanned out to the crowd, totally unresponsive and playing with their laptops.
posted by mikedouglas at 5:35 AM on May 24, 2008

mystical security properties attributed to NAT by so many networking/IT "professionals."

This system is shutting down. Please save all work in progress and log off. Any unsaved changes will be lost. This shutdown was initiated by NT AUTHORITY\SYSTEM

Time before shutdown: 00:00:59
posted by CautionToTheWind at 5:46 AM on May 24, 2008 [1 favorite]

well, it's utterly impossible we could have more than one internet, right?
posted by pyramid termite at 6:25 AM on May 24, 2008

My stint as a product marketing manager in telecom/datacom hardware for 7 years has taught me two things:

1) No one can really predict how networks will evolve

2) Everything takes MUCH longer to evolve than even the most pessimistic prediction
posted by ZenMasterThis at 6:34 AM on May 24, 2008

It is an interesting article, but it is an article that starts from the questionable argument that IPv6 will never happen. Yeah, it has taken a helluva long time to get there, but so has the HDtv transition. If anything, to me, this article shows why NAT won't work.

The one thing the article leaves out is a discussion of ubicom and/or the Internet of things. Once these things become adopted beyond S.Korean planned communities and billionaires homes, won't the hunger to capitalize on this market just necessitate that IPv6 be used?
posted by 8 Bit at 6:41 AM on May 24, 2008

more than one internet, right?

The center must hold.

Given all the stuff that could tear it apart: taxes, legislation, net-neutrality, IPv4, etc..perhaps a perfect storm will make the Internet FUBAR and a new cycle of mom and pop ISPs springing up to battle the big guys will start all over again. It would probably be based on something like WiFi where there is no regulation or need for telco infrastructure.

Yeah I remember the day in 1993 I got a full routed Class B from the NIC, 168.143.0.0/16 - had to do a little white lie on the application form but wasn't much of a problem. No longer have it but occasionally see it in use and think "there's my child", all grown up and scattered to the four corners of the planet.
posted by stbalbach at 6:47 AM on May 24, 2008 [1 favorite]

> This sounds suspiciously like a circular argument to me, but I don't know if there's a specific technical meaning to "IPv6 will never be an alternative to IPv4" that I'm missing.

Woodyatt's been a participant in IETF IPv6 working group(s) for a while; I would imagine he's investigating a scenario rather than sharing his vision of the grim meathook future.

I mostly agree with the scenario, although I think IPv4 address space will ultimately be traded in a speculator's market, similar to how NetSol, and then the regulating body it endowed, ensured that domain name ownership and regulation took the lowest road possible. Americans don't like surrendering control of anything to international regulation, except when they can do all the rulemaking and enforcement themselves.
posted by ardgedee at 7:29 AM on May 24, 2008

What's interesting is that IPv6 has been very well planned from a transition point of view. A very clear path for v4 and v6 in parallel. Ten years of stable spec. Strong vendor support: every consumer OS supports IPv6. Plenty of early adopters, too. And yet IPv6 is still not a real network. In my mind the missing link is routing support; my cheap home router, my ISP. I don't know if that's the only problem, but it's sure stopping me.

Google just launched search on IPv6. You can try it at ipv6.google.com. Here's the DNS entry:

ipv6.google.com.        10776   IN      CNAME   ipv6.l.google.com.
ipv6.l.google.com.      300     IN      AAAA    2001:4860:0:2001::68

posted by Nelson at 7:53 AM on May 24, 2008 [1 favorite]

What a bunch of fearmongering bullshit. The IPv6 transition is alive and well in other parts of the world, it's just that in the U.S, there is a lot of IPv4 address space.
posted by delmoi at 7:56 AM on May 24, 2008

At work they have a saying: "We're planning to retire the VAX in about 18 months". They've been saying that for years. IPv6 is the same thing - I took a great class on networking from Scott Bradner in 1999 - then and now, it's just a couple years away.
posted by These Premises Are Alarmed at 8:04 AM on May 24, 2008

Meh.

For fifteen years, we've been told the doom that would befall us all if we didn't go to IPng, I mean, IPv6.

The real world answer? NAT, CIDR, BGP/IGRP, and cheap CPU/RAM making the large and complicated routing tables that BGP/IGRP generate workable.

Then, the feature set was too compelling.

The real world answer? Port these killer features to IPv4. The only ones I know that hasn't been is stateless autoconfiguration and SEND, and that's because they've turned out to not be killer features in IPv4.

Now, we're being told that the world will end if we don't go to IPv6.

Again.

Call me when 90% of the IPv6 network isn't misconfigured.
posted by eriko at 8:59 AM on May 24, 2008

We just need to drill ANWR for more IPv4 address space.
posted by George_Spiggott at 10:05 AM on May 24, 2008 [1 favorite]

IPv6 is coming.. touch her, feel her, she's cumming.... OMG yes!

v6 has been a talking point for a few years now. This year it was a raise your hand and commit to implementing thing around the .edu circles. A lot of we have to start and make it work. Cali is working hard to get v6 space working. .edu people are working hard. I have v6 on my laptop at home.... you'll get it, it's nice once you get used to it.
posted by zengargoyle at 10:10 AM on May 24, 2008

That governments are being mobilized and billions spent because there is a number shortage is some sort of bizarre commentary on civilization. Imagine what Swift could have done with such a thing! Somewhere, a Pythagorean is laughing at you.
posted by enn at 10:37 AM on May 24, 2008 [3 favorites]

Most of you need to scroll up and click on the link provided by sdodd above. It's an intelligent treatment of what the state of the IPv6 transition is. Nelson, you'll see that IPv6 is not yet completely supported in consumer OS's.
posted by intermod at 11:13 AM on May 24, 2008

More to the point, NAT is already in wide use, and apparently we still need 170 million new IP addresses every year.
posted by furtive at 2:04 PM on May 24, 2008

rokusan wrote: I kinda like being able to memorize, or at least recognize an IPv4 address, myself. I don't look forward to the days of scanning logs full of "fe80::200:f8ff:fe21:67cf" and wondering who the heck that is.

There is this thing..I believe they call it "DNS." Someone once told me it translates numbers to names and vice versa. Perhaps you should look into it.
posted by wierdo at 9:09 PM on May 24, 2008

IPv6 is not yet completely supported in consumer OS's

Hardly surprising given that Every OS Sucks. (Wow, I can't believe that's seven years old.)
posted by PeterMcDermott at 12:46 AM on May 25, 2008

> It is an interesting article, but it is an article that starts from the questionable argument that IPv6 will never happen. Yeah, it has taken a helluva long time to get there, but so has the HDtv transition. If anything, to me, this article shows why NAT won't work.

I believe that's the point of the article. It's taking the idea that IPv6 isn't necessary to an extreme conclusion: IPv4 will rule the world and anarchy will reign supreme.
posted by Monochrome at 10:15 PM on May 26, 2008