Wednesday, June 07, 2006
IPv6 coming to a settopbox near you?
No technology is successful without solving a real problem. No infrastructure survives long without users.
For a long time it looked like IPv6, the next generation of the most important internet protocol (the IP in TCP/IP), was a solution looking for a problem to solve.
To be more precise, the problem was clear: lack of Internet addresses for the world at large (see my analysis a while ago).
Most unclear however, and largely unanswered, was the question whose problem it was. For every economic actor in the Internet world, the move to IPv6 signifies additional expense and complexity. We have enough of that already.
On top of that: the ISPs like to see marketshare before they move, the software vendors and equipment vendors are waiting on the application providers, the application providers are waiting on the users, and the users are waiting on the ISPs and the IPv6 aware applications. A prisoners dilemma because you need the other people to cooperate for you to benefit.
Recently however, Comcast (a large cable company) bit the bullet. After an internal analysis revealed the need for more than 100 million IP addresses in the near future, the business case for IPv6 was made.
A few important lessons out of this case:
Thanks to Jaap Akkerhuis for bringing this to attention.
For a long time it looked like IPv6, the next generation of the most important internet protocol (the IP in TCP/IP), was a solution looking for a problem to solve.
To be more precise, the problem was clear: lack of Internet addresses for the world at large (see my analysis a while ago).
Most unclear however, and largely unanswered, was the question whose problem it was. For every economic actor in the Internet world, the move to IPv6 signifies additional expense and complexity. We have enough of that already.
On top of that: the ISPs like to see marketshare before they move, the software vendors and equipment vendors are waiting on the application providers, the application providers are waiting on the users, and the users are waiting on the ISPs and the IPv6 aware applications. A prisoners dilemma because you need the other people to cooperate for you to benefit.
Recently however, Comcast (a large cable company) bit the bullet. After an internal analysis revealed the need for more than 100 million IP addresses in the near future, the business case for IPv6 was made.
A few important lessons out of this case:
- the traditional approach to address space shortage is using private address space (10.x.x.x addresses). Unfortunately, this fails here, because there are only about 16 million of these.
- to get things going, Comcast is first using IPv6 for the management of the set top boxes and cable modems, not directly for the customers usage.
Thanks to Jaap Akkerhuis for bringing this to attention.
