Short answer: I think it’s students. Here’s why. Google has a great graph of IPv6 usage at google.com/ipv6/statistics.html,* currently measuring about 0.74% IPv6 usage – a woefully small percentage, about 3 IPv6 users per 400 IPv4 users. Below is a screenshot for 15 July. Ignore the red and blue lines and just look at the green one, native IPv6 – clearly now the preferred form of IPv6 connectivity.
See the section under ‘1’? The purple dot is when World IPv6 Day 2011 occurred, and the fall in IPv6 usage after it was widely interpreted as people giving IPv6 a go, then dropping it. Sounds reasonable, till you look at the graph under ‘2’ and ‘3’. In 2009 and 2010 thare are similar fluctuations, yet there were no ‘World IPv6’ events then at all. So why would IPv6 usage change like this each year?
Those with long memories of Usenet will know. Northern hemisphere student summer holidays are from June to September, and in the early 1990s, when the Internet was the sandbox of academia, each September would bring a flush of clueless newbies, intoxicated by their first encounter with this astonishing new world. (That is, until September 1993, when AOL and Compuserve opened up access, and the great Eternal September of the Internet began.)
Today, IPv6 is in a similar position. The major early adopters are academic and research networks, so for the time being, IPv6 usage will show variations following the northern academic year, and students will drive the growth of IPv6. See The Pirate Bay Goes IPv6 and Slow IPv4 Spurs IPv6 Alternative for some insight into why.
Why does this matter? For a start, the fall in IPv6 usage seen last year after World IPv6 Day was not due to disinterest, it was due to students leaving their IPv6-connected universities for the summer. A close look at the graph marked ‘4’ shows the same decrease starting after this year’s World IPv6 Launch. Gloom-and-doom merchants may again say this is a failure of IPv6, but I suspect there’ll be a burst of IPv6 usage in early September 2012, and growth will continue. Let’s see what happens …
More importantly, an observer in 1992 of the newbie-infested networks would have been amazed to be told that within five years the Internet would have exploded into essential global infrastructure. Even if IPv6 grows the same way, many might say so what? It’s just a protocol to help the Internet do what it already does so well. But that observer in 1992 would also not have known that the World Wide Web – just another protocol – was about to change everything all over again. Can IPv6 bring about that kind of game-changer? I don’t know, but I do know that, like IPv6 addresses, innovation on the Internet is (almost) infinite.