Good query.  I too was wondering what PN are planning, as I read that some ISPs have been offering IPv6 addresses to customers.  I also wonder how the internet 'community' are going to offer a smooth transition. 
I had been wondering if NAT might have been an option so an external IPv6 address might be used for people with dynamic IP addresses using websites which are actually on IPv6 hosts, or whether there needs to be a 'major switchover' day.
I know Windows 7 is 'IPv6 ready' but there must be millions of households with PCs (WIn 98 etc) and kit which don't yet have suitable software in them (and some hardware might need to be replaced anyway, as the work to alter them would cost as much as a new device - the sub 15 quid 5 or 8 port ethernet switch, for example, would cost a few quid in postage to send away and return, and a new circuit board, or even changing the firmware, might cost for parts and labour labour, so a new box may be sensible option).

Quote from: WebDude

Good query.  I too was wondering what PN are planning, as I read that some ISPs have been offering IPv6 addresses to customers.  I also wonder how the internet 'community' are going to offer a smooth transition.

I'm currently playing with IPv6 tunnelling from HE.

Quote

I had been wondering if NAT might have been an option so an external IPv6 address might be used for people with dynamic IP addresses using websites which are actually on IPv6 hosts, or whether there needs to be a 'major switchover' day.

Not totally clear what you are getting at here.  But I will answer what I think you are asking.  IPv4 services can work alongside IPv6 services, so there is no need for anything to be done in terms of either keeping existing IPv4 services going, or a major switch over day, IPv4, even once its reached its address limit can continue to work as it is now for many years.

Quote

I know Windows 7 is 'IPv6 ready' but there must be millions of households with PCs (WIn 98 etc) and kit which don't yet have suitable software in them (and some hardware might need to be replaced anyway, as the work to alter them would cost as much as a new device - the sub 15 quid 5 or 8 port ethernet switch, for example, would cost a few quid in postage to send away and return, and a new circuit board, or even changing the firmware, might cost for parts and labour labour, so a new box may be sensible option).

Most home network equipment and older PC's with older operating systems will need to be upgraded or replaced to work with IPv6, but then again, on your home network, there is no reason why you cannot continue to use the private IPv4 address space and allow your home network to work as it always did.
It is however when people start to launch new services post IPv4 address exhaustion, at that point there will suddenly be demand for IPv6 solutions, some homebrew ADSL modems (OpenWRT for instance) support, or are beginning to support IPv6, but the likes of Plusnet, BT, Netgear, Linksys (Cisco by another name) need to be getting solutions and roll out plans in place now, a full IPv6 infrastructure will take years to build and certain companies and organisations are already ahead of the game and ready, but others (and it seems Plusnet is in this category) are not paying it any attention, yet as best as I can see and understand the problem is months away from blowing up in peoples faces.

Hmmm, interesting thread, in so much as it says a lot about Plusnets attitude.  For me its a very head in sand approach at the moment, Ben is correct in saying that just because no more IP's can be allocated to ISP doesn't mean the ISP will not have spare addresses, but of course its going to get really controversial as if a new entrant wants to come to the market they will be bared from entry as they wont have any IPv4 addresses, I can see perhaps 3 or 4 years from now the monopolies (perhaps OFCOM) people ripping IPv4 addresses away from the current holders and allocating them to others.  IPv4 isn't going away any time soon, but ISP's, service providers and equipment suppliers need to get real, selling IPv4 only services and equipment, and rolling out IPv4 only solutions at this late stage is at best silly (you listening BT?) or at worst going to lay them open to law suits, in US at least I can already see in my minds eye the class actions taken out by people who bought Netgear, Linksys etc equipment and find out that within a few months of purchase, and with the full knowledge of the manufacturer, the device is now a bar to the use of the internet not an enabler.
IPv6 roll out for an ISP is not a simple issue, IPv6 is not just a larger address space (i.e. its not just a larger IPv4), routing works somewhat differently, and there are difficult (as in not obvious) choices to be made by ISP's as to how they integrate IPv4 and IPv6 traffic and address spaces, peering agreements will need to be relooked at as will basic peering arrangements...
Then the next issue the ISP's will be forced to deal with, is consumer equipment problems, very few consumer modems, wireless points etc offer IPv6, and many will never be able to, they just don't have the juice needed for the extra code needed to support both IPv6 and IPv4, which is compounded many times over by the fact that most people have never upgraded the firmware in their modem/router and would be quite intimidated by the idea of doing so.
The sooner an ISP looks at these issues the sooner that they can be sorted, firstly by doing obvious things such as only sending customer IPv6 routers, and my putting pressure on the likes of Netgear etc to get their act together on IPv6 long before there is any consumer clamor.  An ISP tackling this early also has time and not so great pressure to sort out the routing, peering and business agreements surround IPv6 traffic and they can start to iron out their own internal systems by running small consumer IPv6 trials, all again whilst IPv4 still does everything, i,e. there is still a window during which ISP's can rollout IPv6 under no pressure, but instead we appear to waiting (we as in the vast majority of the Internet) until the pressure is on, when it will be costing $$$$ when it goes wrong.
*sigh*
As I say, head in sand stuff here.

the technology behind the 21CN  network is kinda familiar to me, it is at least virtually all IP, even if its industrial scale and class, and I've worked at Internet companies and worked on large IP network roll outs, so whilst it seems foolish that the 21CN doesn't support IPv6 out the box (so to speak) I can at least get my head around how that would be upgraded to work with IPv6 and can see that for the most part it will be a fairly easy task to do, so long as their staff are properly trained (OMG this is BT were talking about  there staff even as ADSL took off still couldn't grasp, explain or for the most part deliver consumer ISDN, properly trained?  BT?  Not a chance) and the commercial/(and peering agreements) are in place.
The technology of the 20CN escapes me, as I suspect it does many, my understanding is it a many layered beast with some technology dating back still to early days of telephones and equipment dating back to the 1980's and then new technology and equipment layered on top.  I'm sure that IPv6 could be delivered on 20CN but the cost and headache in doing so maybe the final push BT needed to upgrade all to the 21CN network.

Article and some FAQ here from an ISPs point of view:
http://www.aaisp.co.uk/kb-broadband-ipv6.html

Quote

IPv4 address space looks to run out some time in 2011 and even that may be optimistic.

Quote from: fourfourdevon

In a year to 2 years IPv6 will be mandatory stuff, not because any regulator says so, but because if you want to expand the internet beyond where it is in 2 years time, IPv6 will be the only real option.

Or, unfortunately, carrier-grade NAT... 
Mathew

Quote from: MJN

Or, unfortunately, carrier-grade NAT...

As practiced now by those mobile networks which have two NAT layers.

Quote from: MrC

As practiced now by those mobile networks which have two NAT layers.

Indeed, and it can be a painful situation to be in for all concerned.
Worse still, some of those networks have decided that they needn't use private address space on their side of the NAT, opting instead to use public address space that they think aren't in use by those it's been assigned to... Not a problem until the rightful owner does start using it on the Internet at which point noone from behind the NAT can reach it. (Been there, got the t-shirt for that one).
Another vote here for doing things properly, with IPv6!
Mathew
P.S. Can I put a plug in here for the World's First IPv6 Cat Feeder?!

Hi,
The IPv6 project is big, probably one of the biggest projects we'll ever do. At the moment I can't answer every single question about what's going to happen because there are still some outstanding questions.
Some things we know, IANA will run out of IPv4 addresses in about a year and RIPE will run out a year or so after that, at which point there will be no more IP addresses that they can give out. Of course that's how many are being allocated, there will be IPs that will be unused at that point so will be a while before they run out.
Of course we know that IPv4 will run out at some point so are preparing for it. We're already ensuring that core network kit is capable of supporting IPv6 and we're currently talking to the CPE manufacturers about what their plans are for IPv6 so that the next routers we supply are IPv6 capable either out the box or via a firmware update.
We're also talking to BT about what their plans are for IPv6 and of course our other partners and peers. We're also looking at how we currently use our IPv4 addresses and seeing how they could conserved and if they can be utilised.
Like I say I can't tell you exactly what's going to happen because at the moment not everything has been decided but a lot of work is currently going off in the background.

Quote from: MJN

Or, unfortunately, carrier-grade NAT

Yeah, that sucks, NAT behind NAT just gets a real pain and breaks a great deal of applications.  However it is inevitable that it will happen, and it will happen more and more.  For a long time, even if every UK ISP has a IPv6 network and all UK consumers have IPv6 routers there will services, servers, locations etc that are only on IPv4.  Yet new services, locations and servers will need to be added, and they will want IPv4 addresses as well.
The upshot will be that for quite some time, IPv4 will continue to grow via the use of carrier NAT on the consumer side and by VIP (virtual IP, imagine a proxy in reverse, one IP address that people talk to, that then distributes those "conversations" to a range of servers, the servers themselves will increasingly not have an IPv4 public address at all, if you will specialised NAT for servers) on the server side.  This will start to cause any number of problems for IPv4 applications and will I'm sure end up causing routing problems and for some segments of the internet to be totally unaddressable by other segments of the internet.
Thankfully, *if* the ISP's have got their acts together (and most don't appear to have at all) this will drive more and more application traffic into the IPv6 sphere where addressability and availability of addresses a none issue.
Reading Dave's contribution to this thread it would seem that Plusnet have decided that IPv6 needs to be tackled, but are still in the thinking and talking phase, there seems very little action,  hope that action starts soon, as Dave says its a big project and I know its not simple either, so getting the internal core network so that it actually does IPv6 would be a huge start and a great learning/test bed for the engineers, and since apparenty the kit is already capable of IPv6 the only ingredient missing here is a willingness to act.  Once that is done I'm certain that Plusnet could find some peers (hell I know my many IPv6 ready ISP complain about the lack of peers) and once they've done that I'm certain they could find a few dozen customers willing to get IPv6 up and running on their home networks (err.. Hello!).  Actions speak louder than words.

Quote from: dave

The IPv6 project is big, probably one of the biggest projects we'll ever do. At the moment I can't answer every single question about what's going to happen because there are still some outstanding questions.

It's great news to hear that moves are being made for the provision of IPv6 to end users!  What would also be great is to have periodic updates on your progress and to keep us informed; is this something you're also planning to do?

Getting IPV6 to end users isn't all that simple (I wish it were!).  There are not many routers that support IPV6 out of the box and if you want one expect to pay around £100.  As far as I'm aware, Netgear offer *no* IPV6 enabled routers.  We could end up in a situation where you have to have a separate router and modem combo to get IPV6.  The modem plugs in to the router, the router supports IPV6 and does the IPV6.... stuff.
It is possible to get IPV6 over an IPV4 network, I believe this occurs by enclosing the IPV6 packets in IPV4 data and connecting to an IPV6 provider (like sixx or hurricane electric), but then you get into the "what the heck is a /128? Why do I want a /64?  /48?  What next, /-1+pi?!?!?".
I am IPV6 enabled and ready to go (Cisco 877W) and there is an ISP in the UK which does provide IPV6 addressing but it's not really worth ISP's time to invest in IPV6 currently, most of the net is IPV4 so having IPV6 would provide a service very few people would use or require, when the big sites start to enable IPV6 (youtube, facebook, download sites, all the popular ones) then ISP's can start to look towards IPV6, but then it becomes a massive problem on the hardware front as well as training the staff to understand IPV6 problems, people on older OS's that don't support IPV6, OS's which are not very good at IPV6 and so on and so forth.
Then we get into the trials stage, PlusNet won't just release IPV6 into the wild, they would trial it first, but to trial it you would need people who have IPV6 enabled routers (I can't see PN sending out lots of expensive Cisco gear!) but there is no demand for IPV6 routers so few people have them and yeah, it's catch 22 at the end of the day.  Until there is a push by everyone to move off IPV6 with a set shutdown date (i.e. a date where IPV4 gets switched off and becomes unusable) then there will be no rush for IPV6 to move along.