Perry Lorier wrote:
Andy Linton wrote:
That kind of implies that if all those Windows users were to turn on the relevant hooks in XP then it would all burst into life. Perhaps Vista will have it on by default and everything will just start working. (:-)
Vista does apparently enable v6 by default. If it can't get a native v6 address but it has a globally routable (ie non rfc1918, etc) v4 address it will use 6to4 and tunnel IPv6 over IPv4. If it can't get a native v4 or v6 address it will use Teredo and tunnel IPv6 over UDP over IPv4 from behind NAT to get a realworld IPv6 address.
This suggests that when(if?) the majority of people are using Vista then they will have at least one real world, globally scoped IPv6 address on their machine, which applications can (if they want) use.
This brings me to ask the question of, "What will happen when this is deployed to the customers who complain about lousy performance?". If you're having to tunnel to some random remote gateway on the Internet, which is 200ms away from you, and 200ms away from the content, suddenly your performance becomes [even] worse. Could we see ISP helpdesks suggesting disabling v6? Or would we see this as an encouraging sign and more 6to4 and Teredo gateways deployed? I know which my money (20c, Australian) is sitting on right now ;).
So we're back to "Who pays?" again! For training and new hardware.
Who pays for v4 training and hardware at the moment? Why won't they pay for v6 training and hardware if/when IPv6 becomes reality?
The customers - either directly or indirectly. Until the customers push it to be a reality [by paying], there isn't a huge amount of drive to learn or build it. As Steve and Dean have pointed out, resource goes where the profit is. Perhaps a new IPv6 guide for managers and telcos? 1. Deploy v6. 2. ???? <-- probably what the customers will do! 3. Profit! aj. -- ipv4 and atm. yeah.