TL>> My long term dual stack future is assured at this stage,
MvV>> For how long? 10 years? After that it probobly won't matter any MvV>> more.
TL> Given the low penetration of IPv6 here, 10 years is not out of the TL> question (I'd rate it better than 50/50).
I am a bit more optimistic. I expect that onc IPv6 becomes the dominant protocol, it will go very quick. In Belgium IPv6 already is the dominant protocol.
MvV>> My choice is limited. My present ISP has a quasy monopoly in my MvV>> situation. They offer a good deal regarding price speed and [..] TL> Here, the situation is different. Few ISPs actually own access TL> infrastructure, and with the new national network rolling out, pretty TL> much all will be dependent on this network to reach their customers.
I think that is a good development. Here the owner(s) of the old copper originally designed for POTS are by law required to allow competitors to make use of the infrastructure. But oddly enough that does not apply to the coax cable originally designed for broadcasting TV. The owner of the coax has a monopoly on it.
TL> Still a good question. :) Not one I have to consider for quite a TL> while, though I have proven that as long as I have access to a public TL> IPv4 from _somewhere_, a tunnel will work, and I'm sure that will hold TL> even if I have to run my own NAT on the far end of the VPN.
So you are planning on supporting incoming IPv4 for as long as you can. If need be at extra cost.
I am going the other way. I plan on fasing out IPv4 as soon as practical. Dual stack and DS-Lite are transition mechanisms. One can not switch from IPv4 to IPv6 overnight, because that would require all of the word wide internet to throw the switch at the same time. That is not going to work, so we need a period of overlap where both IPv4 and IPv6 are supported. But I am not a member of the "more is better club". Supporting two protocols in parallel increases the chance of bugs and the effort required to trace and fix them. Also I would gladly get rid of NAT. The sooner I can get rid of IPv4 the better I say. I encourage the use of IPv6. But when the IPv6 train really gets on steam, the next step would be to discourage the use of IPv4.
I have been running IPv6 via 6to4 tunnels for over half a decade. Technically the reverse, IPv4 via a 4to6 tunnel should work just as well. If need be, I will go that way. But that is plan C. Plan A is to have dual stack until my provider forces me to DS-Lite. My hope is that when that happens, I no longer need a public IPv4 address. Plan B is to switch provider.
