back to article IPv6 now faster than IPv4 when visiting 20% of top websites – and just as fast for the rest

Accessing websites via IPv6 is not only comparable in speed to IPv4, but is actually faster when visiting one in five of the world's most popular sites, according to German researchers. In a new paper, Vaibhav Bajpai and Jürgen Schönwälder from the University of Bremen looked at the response times for the internet's top 10,000 …

  1. ZeroSum

    And what about BT and Virgin?

    What's the story with BT's deployment? Troubles with the IPv6 support in their home hubs perhaps. They were supposed to be at 50% deployment by now with it completed by the end of the year. LOL. That was never believable given their past performance. In fairness it should not be difficult for them to fix if they just put a little resource into it.

    Meanwhile Sky Broadband are at 70% deployment.

    1. Tom 7 Silver badge

      Re: And what about BT and Virgin?

      "IPv6 will be disabled on your BT Home Hub and BT Broadband Network until supported by future services"

      according to my router.

    2. AndrueC Silver badge
      Meh

      Re: And what about BT and Virgin?

      Some of the smaller ISPs have been at 100% deployment for a couple of years now. AAISP have as has IDNet.

      Meanwhile Plusnet ran a couple of trials back around 2009 and seem to have given up. Oh well.

      1. ZeroSum

        Re: And what about BT and Virgin?

        > Meanwhile Plusnet ran a couple of trials back around 2009 and seem to have given up. Oh well.

        Whoever did that at Plusnet is probably long gone.

        There's a long tail of small ISPs. Some technically competent like AAISP and majority that are controlled by the bean counters.

      2. John Sager

        Re: And what about BT and Virgin?

        Some of the smaller ISPs have been at 100% deployment for a couple of years now. AAISP have as has IDNet.

        I used to be with an Entanet reseller years ago, and got connected to Entanet's experimental service, and then after a hiatus, to their main service (on a different prefix). However that eventually failed when some lash-up kit they used for connections via BT 20C networks failed. I got fed up of waiting for them to fix it & moved to AAISP a couple of years ago. I very occasionally see v6 outages - there was one yesterday for a while - which I notice when, particularly, fonts.googleapis.com hangs (lots of websites seem to use this).

        Naturally I'm all dual-stack here for all my hosts except for backward vendors such as the TV kit. It would be good to go v6 only but that needs a 6-to-4 proxy service somewhere for all the v4-only services out there that will never die, and I can't do it until all the v4-only kit here goes to recycling.

        1. AndrueC Silver badge
          Happy

          Re: And what about BT and Virgin?

          I'm all dual-stack here

          Me too, thanks to being with IDNet for a year. I went through three or four routers(*) until I found one that could handle dual-stack and was stable. Ended up with a Billion 6300. I even worked out how to get my mail server to play over IPv6. Sadly for the past couple of years my router has only got an IPv4 address to play with but if/when PlusNet ever wake up I'll be ready :)

          (*)One of which, a TP-Link, claimed to be IPv6 ready but didn't support a dual-stack configuration. Not terribly useful :-/

    3. Lee D Silver badge

      Re: And what about BT and Virgin?

      The Virgin one, especially, pisses me off.

      They're always holding conferences about it and saying how it's coming soon and nothing appears to the ordinary people.

      I'm on a Virgin leased line at work - same. Not a single mention of it.

      BT, I can understand. They're always behind the times. But Virgin have no excuse - DOCSIS 3 standardisation basically requires IPv6 compatibility. I get that even their own hubs might not be able to do anything, but there's no reason they can't pass the traffic through when in modem mode.

      But without 6rd or DHCPv6 or similar, there's no way to get a native IPv6 and we're all still stuck on tunnel-based IPv6 which is like saying "Yeah, sure, you can get on the Internet - just let this other random third-party (that you're not paying and is usually based in the US) sniff all your traffic".

      1. ZeroSum

        Re: And what about BT and Virgin?

        > The Virgin one, especially, pisses me off.

        Badger them to do it dual-stack instead of using DS-Lite. If necessary they should buy more IPv4 from poor countries. Most of Liberty Global have deployed IPv6 using DS-Lite and it is terrible for power users that are still dependent on having a public IPv4 address on their WAN link so they can port forward.

      2. Grifter

        Re: And what about BT and Virgin?

        Sad to say but your traffic is most likely sniffed by someone somewhere, whether you use a tunnel or not, so don't let that argument stand in your way.

  2. Anonymous Coward
    Anonymous Coward

    We didn't run out of ipv4

    They're just horribly inefficiently allocated.

    A full quarter of the address space is assigned to, but hardly utilized by, HP, Ford, Xerox, Bell, Apple, MIT, DoD, MOD, IBM, Boeing, DuPoint, Halliburton, AT&T, Prudential, US Army, US postal service.

    AND the same mistakes are being made all over again. Massive chunks of ipv6 are being handed out because "the address space is just so damn huge". If you assign blocks of a quintillion addresses at a time you're going to run out.

    Don't get me wrong, getting rid of NAT is good, and I operate an ipv6-only website which really helps to keep the peasant scum out.

    1. ZeroSum

      Re: We didn't run out of ipv4

      Time to stop beating that drum.

      We're going to need a lot more than 3.5 billion independent Internet connections. Every IPv4 connection that isn't behind a NAPT needs a public IPv4 address. The only way the IPv4 Internet scales is ever more NAT and middleware.

      IPv6 is being allocated in a sensible way. There are 18 billion billion /64s and only a small amount of that is being allocated from.

      http://www.iana.org/assignments/ipv6-unicast-address-assignments/ipv6-unicast-address-assignments.xhtml

      1. Anonymous Coward
        Anonymous Coward

        Re: We didn't run out of ipv4

        Well, I can agree in part… do I really need a whole /64 to run a home network? As it happens, I've got use of a /56 from my ISP, and I do use a few /64s out of it.

        Those /64s could easily be /48s without issues… or using DHCPv6, even smaller. But /64 is the smallest an IPv6 subnet will go, so /64 it is for those subnets.

        1. Dwarf Silver badge

          Re: We didn't run out of ipv4

          @stuart a /48 is bigger than a /64 in the same way that a /8 is bigger than a /16 in IPv4 :-)

          The /64 is there to optimise routing, which means faster connections.

          Ask yourself this - given any "normal" home network, does anyone actually need an IPv4 /24 netblock ? do you have >200 devices at home ?

          The same logic applies in IPv6 - there are some spare addresses in your local subnet - so what ?

          The benefit is that auto-configuration relies on this standard network length, your devices can auto-configure without having to worry about DHCP and the like, so its easier to set up.

          1. Down not across Silver badge

            Re: We didn't run out of ipv4

            Ask yourself this - given any "normal" home network, does anyone actually need an IPv4 /24 netblock ? do you have >200 devices at home ?

            Yes, I do1. And no, I don't want them all addressable from the internet and am quite happy to subnet them (RFC1918) and NAT and port forward as needed.

            So while I could use a /24 netblock, I don't actually need one. Realistically /28 or /29 would be perfectly fine on IPv4.

            1 Yes, small number are virtual interfaces / aliases

          2. Anonymous Coward
            Anonymous Coward

            Re: We didn't run out of ipv4

            @stuart a /48 is bigger than a /64 in the same way that a /8 is bigger than a /16 in IPv4 :-)

            Yeah, sorry, you're quite right. I meant to go the other way. /72s?

        2. ZeroSum

          Re: We didn't run out of ipv4

          Since SLAAC uses /64 and modern hosts randomly pick an address to use a pure IPv6 address scan of the /64 becomes too costly to be practical. They have to get you to visit a honey pot site to discover what your address is.

        3. oldcoder

          Re: We didn't run out of ipv4

          It allows you to use your MAC address (48 bit) as the address on your /64 network...

          No need for DHCPv6.

          Yes, it would be nice if you subnet your /64 network - but not mandatory.

          1. Preston Munchensonton
            Boffin

            Re: We didn't run out of ipv4

            It allows you to use your MAC address (48 bit) as the address on your /64 network...

            No need for DHCPv6.

            Yes, it would be nice if you subnet your /64 network - but not mandatory.

            Sigh...you don't use your MAC address as your IPv6 address. It can form the basis of an IPv6 address using EUI-64. Definitely not the same thing.

            Regarding DHCP, are you planning to just continue to run your DHCPv4 servers? If not, every host will not receive autoconfiguration for all of those lovely DHCP options not covered by SLAAC (e.g. DNS, WPAD, PXEboot, etc.). DHCPv6 fills in that gap if the plan requires completely dismantling the IPv4 infrastructure.

            Anyone who thinks we'll be completely off IPv4 anytime in the next 40 years is trying to sell you something.

        4. Yes Me Silver badge

          Re: We didn't run out of ipv4

          [Fact check: a /48 is a bigger network than a /64.]

          To run a home network of the future (with multiple subnets) you need a /56 or a /48. But given that 15 trillion /48s are readily available, this really isn't a problem or a waste. The /64 boundary is explained in RFC7421.

      2. Anonymous Coward
        Anonymous Coward

        Re: We didn't run out of ipv4

        >We're going to need a lot more than 3.5 billion independent Internet connections.

        Your toaster is going to talk back to LG corporate headquarters and you are going to pay for the inet access bitch else how can they check if your subscription is up to date. And don't touch that DRM asshole. First they came for my game console ...

        1. ZeroSum

          Re: We didn't run out of ipv4

          > Your toaster is going to talk back to LG corporate headquarters

          For there to be more than 3.5 billion internet connections in the world that don't sit behind a NAT the larger address space of IPv6 is needed.

          The important scaling with IPv6 is not the number of devices sitting on a customer home LAN. It is the number of possible customer LANs that don't have the restrictions of NAT.

          1. Anonymous Coward
            Anonymous Coward

            Re: We didn't run out of ipv4

            >The important scaling with IPv6 is not the number of devices sitting on a customer home LAN.

            But it sure is nice for the shareholders if every one of them is publicly addressable.

            > It is the number of possible customer LANs

            Yep as if there isn't enough motherfsckers in the world already. Have a strong feeling its not going to be the internet (IPv6 or not) that won't scale. Fresh water anyone but I digress badly.

        2. ZeroSum

          Re: We didn't run out of ipv4

          > Your toaster is going to talk back to LG corporate headquarters and you are going to pay for the inet access

          Your toaster would be able to make an outward call home through an IPv4 NAPT in the residential home gateway, so the corporate overlords will know about it either way.

          IPv6 eliminates the need for the NAPT and allows for transparent end-to-end connectivity. That will allow applications that don't require centralised control.

          1. Anonymous Coward
            Anonymous Coward

            Re: We didn't run out of ipv4

            NAT and firewalling are two completely different things. You can have firewalling without NAT.

            > IPv6 eliminates the need for the NAPT and allows for transparent end-to-end connectivity. That will allow applications that don't require centralised control.

            Except, would you really configure your firewall/gateway to allow all inbound traffic to all devices on your home network?? I doubt it.

            The IoT is still going to use "meet in the middle" servers, with or without NAT.

            1. Number6

              Re: We didn't run out of ipv4

              The IoT is still going to use "meet in the middle" servers, with or without NAT.

              What I want is some sort of IoT standard such that I can run my own server, point all my own IoT things to it and then have a good restricted access to that from the outside so I can get my phone or laptop to call home. That way it all stays within my firewall, under my control and (bar vulnerabilities in the server), keeps all my data safe. Of course, it also keeps it safe from all the companies out there that want to collect my personal information, which is why they will probably never conform to a common protocol.

            2. ZeroSum

              Re: We didn't run out of ipv4

              > Except, would you really configure your firewall/gateway to allow all inbound traffic to all devices on your home network??

              It should support the IPsec and Internet Key Exchange (IKE) recommendations of RFC 6092.

      3. Anonymous Coward
        Anonymous Coward

        Re: We didn't run out of ipv4

        > There are 18 billion billion /64s and only a small amount of that is being allocated from.

        Note that several years ago, France Telecom got assigned a /19 block of IPv6 addresses. That is 1/65,536 of the *total* unicast IPv6 address space (the first three bits are fixed at 001).

        If ISPs keep burning it like that, then we will indeed need to go to a completely new address plan shortly after IPv6 actually takes off.

        The problem is SLAAC. Nobody wanted a replacement for DHCP; it works just fine. And although using 64 bits containing your MAC address is a great idea for link local addresses, it's stupid for global addresses; it then led to the whole new requirement for "privacy addresses".

        If this had been done right, every end user would have got a /96 assignment (with 32 bits to assign themselves - as big as the whole Internet is today!) and used DHCPv6.

        DHCPv6 *is* available and mostly works fine - although Android still doesn't implement it. So apart from Android phones, you can simply turn off SLAAC everywhere.

        1. ZeroSum

          Re: We didn't run out of ipv4

          > Note that several years ago, France Telecom got assigned a /19 block of IPv6 addresses. That is 1/65,536 of the *total* unicast IPv6 address space (the first three bits are fixed at 001).

          That is only the currently delegated by IANA 2000::/3 prefix.

          There's nothing to stop 4000::/3, 6000::/3 being delegated in future.

          I will admit that France Telecom should not have been assigned a prefix as large as a /19. That is ridiculously big.

        2. asdf Silver badge

          Re: We didn't run out of ipv4

          >. So apart from Android phones (devices?),

          Too lazy to look but that my guess that is probably at least close to the majority of things connected to the internet these days.

    2. bombastic bob Silver badge
      Mushroom

      Re: We didn't run out of ipv4

      "getting rid of NAT is good, and I operate an ipv6-only website which really helps to keep the peasant scum out."

      the common misconception here is that NAT is bad, though there ARE cetain security advantages to using it.

      First of all: do you REALLY want publically accessible IP addresses on EVERY device you own? Isn't it a good idea to have your PCs and other things at home behind some kind of NAT-based firewall? It's a fair bet that NONE of your devices that support IPv6 are properly firewalled for it. This goes TRIPLE for Windows. Try a 'netstat -an' some time, and see what's listening... and other machines, ALSO listening on the SAME ports. It should scare you a bit, because NEARLY EVERY IPv6 address is publically visible and routable, and discoverable whenever you 'hit' a web site with an IPv6 request.

      I've personally been using IPv6 for a few years now, using a tunnel through he.net . My ISP doesn't support IPv6 and probably never will. No biggee, it works fine this way. HOWEVER, after early experiments, I determined that windows boxen that have IPv6 configured are DANGEROUS security craters waiting for some 0-day that exploits the "somewhat well known" listening ports. So I scan all of the windows boxen, and add these ports specifically to my (FreeBSD-based) firewall and router box. But I also see the pathetic support for IPv6 that exists on typical NAT-based routers, particularly wifi, particularly older ones.

      Sadly, I think that widespread adoption of IPv6 would yield more virus outbreaks than ever before. Too bad, because it SHOULD be "the standard" by now. Perhaps if we can convince Micro-shaft that it needs to STOP IT with the listening ports on "::" and "0.0.0.0" then security *might* improve. If they MUST listen for connections on these 'somewhat well known' ports, they need to listen on LOCALHOST ONLY, and *NOT* every stinking IP address (and IPv6 address) you've configured. And that just about sums it up.

      1. oldcoder

        Re: We didn't run out of ipv4

        Nothing prevents your firewall/router from blocking devices you don't want public. The problem right now is that those devices STILL get given access to the internet, and that still puts them on the net - and you still don't get privacy; even with NAT.

    3. John Sager

      Re: We didn't run out of ipv4

      Don't get me wrong, getting rid of NAT is good, and I operate an ipv6-only website which really helps to keep the peasant scum out.

      True, for a while. I've just looked at my firewall logs going back about 3 weeks. In that time I've had about 40k v4 'door knockers' that my firewall dropped. The v6 equivalent is essentially zero except for a few odd probes from pnap.net which look like attempts to measure performance. This is on a home network prefix with no outward-facing servers. That won't last but I think we have a few years before it gets bad.

  3. Dwarf Silver badge

    Time to learn

    <soap box>

    Not surprising, given that IPv6 is the preferred protocol on most OS's over IPv4, so it will try that first.

    I've been running IPv6 for about 4 years now. IPv6 isn't hard. Only layer 3 changed and the dependency on Ethernet was removed (no ARP)

    Go get a free tunnel broker (Hurricane Electric or SixXS), they will give you a bucket load of IPv6 addresses to play with (a /64 AND a /48). Hook up one of those spare Raspberry Pi's that are kicking around and you could be running it within the hour.

    Just make sure that you configure your IPv6 firewall correctly (on each device) as there is no need for NAT now. The good tutorials cover this off.

    There are loads of tutorials out there, Google for ipv6 raspberry pi hurricane electric (or sixxs)

    Once you've learned IPv6, then when your ISP does eventually get around to waking up and smelling the coffee, then its an easy change over, you just change the prefix on the Pi to the ISP's one. Alternatively, if your router can do native IPv6, then configure the router and power off the Pi

    In the mean time, you are already skilled up and its not a bad skill for a CV either !

    </soap box>

    1. Nate Amsden Silver badge

      Re: Time to learn

      one of my ISPs is Hurricane electric (co-lo in fremont facility just a 1U server though I have a 3rd of a rack, power limits prevent me from using more though). They asked me if I wanted IPv6. I said...NOPE. don't need it. They did give me some IPv4 IPs, I think I got 6 or something (running vSphere on my server with a few VMs on it, also running an openbsd firewall inline in front of the VMs which does NAT for other VMs that don't have external IPs). I do like the 100meg unlimited connection I have with HE though, I proxy most of my HTTP/HTTPS traffic through the colo.

      I see what you say about IPv6, but if I am going to play around with things I'll go play Fallout 4 or something, playing with IPv6 "just because" doesn't sound remotely fun or interesting to me.

      1. Anonymous Coward
        Anonymous Coward

        Re: Time to learn

        >IPv6 "just because" doesn't sound remotely fun or interesting to me.

        Necessity is the mother of invention. I became an armchair semi expert on *nix networking and QOS on my home routers due to household downloads fscking up my gaming latency. As you say IPv6 buys me nothing IPv4 doesn't for home use (except arguably less privacy).

    2. FIA

      Re: Time to learn

      "Just make sure that you configure your IPv6 firewall correctly (on each device) as there is no need for NAT now. The good tutorials cover this off."

      I thought one of the advantages of IPv6 was that it made NAT actually useful? I.e. you have your internal network on a private /64 subnet then NAT bidirectionally to your external /64 subnet. This means changing IP is easy but you can still get the advantages of uniquely addressable machines.

      1. SImon Hobson Silver badge

        Re: Time to learn

        I.e. you have your internal network on a private /64 subnet then NAT bidirectionally to your external /64 subnet.

        That would actually be Prefix Translation rather than address translation, but unfortunately (AIUI) that got kicked out as "not needed" quite early one. I can see many reasons why prefix translation would be useful (especially if standardised and with a standard way of devices getting the translation info from the device doing the translation).

        This means changing IP is easy but you can still get the advantages of uniquely addressable machines.

        Not just prefix changes, but also multihoming would be easier. Without it, it's easy to multihome a device (just give it an address in each prefix) - but you then have to put the routing decisions (which address they use for outbound connections) on each device instead of being able to manage it in one place.

        This impacts on load balancing (each individual host doesn't know what the combined traffic looks like at the router), and also on failover time in the event of service loss.

        1. Neil Alexander

          Re: Time to learn

          "That would actually be Prefix Translation rather than address translation, but unfortunately (AIUI) that got kicked out as "not needed" quite early one."

          Despite that, you can NETMAP quite easily using netfilter6 to translate prefixes with minimal effort on Linux. In fact, this is exactly what I do on my home network with my EdgeRouter X, which includes these modules out-of-the-box.

    3. bartsmit

      Re: Time to learn

      Actually SiXXS no longer issues new tunnels as a way of exerting pressure on ISP's to offer dual stack to domestic subscribers. https://www.sixxs.net/signup/

    4. Vic

      Re: Time to learn

      Go get a free tunnel broker (Hurricane Electric or SixXS), they will give you a bucket load of IPv6 addresses to play with

      SixXS won't. They sent me a mail, including the following snippet :-

      We are now fully stopping accepting signups and tunnel & subnet requests.

      That was back in April, and they haven't told me any different since...

      Vic.

  4. Anonymous Coward
    Anonymous Coward

    Ironically it will probably be the more tech savy (the more luddite ones anyway) to be the last people on IPv4 only as everyday people's cable modems and router will have IPv6 enabled by default it looks like. I know I am sure to disable IPv6 everywhere I can (computer, router, etc) because #1 my ISP still gives me IPv4 only and #2 I don't want to worry about the twice the attack surface and twice the firewall rules and tc settings at home.

    1. Anonymous Coward
      Anonymous Coward

      And yes I am aware I can run IPv6 on my lan side irregardless if ISP doesn't give IPv6 but I don't and with my lovely 172.x.x.x IPs I can keep it that way forever you IPv6 lovers. Take your FE80:: garbage and stick it where the sun don't shine lol.

      1. Anonymous Coward
        Anonymous Coward

        now just trollin

        NAT 4 life. IPv6 will be perfect when we are all forced to have IoT shit and everything as a subscription in our house.

        1. Anonymous Coward
          Anonymous Coward

          Re: now just trollin

          You write as if NAT doesn't exist on IPv6.

          -t, --table table

          This option specifies the packet matching table which the command should operate on. If the kernel is configured with automatic module loading, an attempt will be made to load the appropriate module for that table if it is not already there.

          The tables are as follows:

          filter:

          This is the default table (if no -t option is passed). It contains the built-in chains INPUT (for packets destined to local sockets), FORWARD (for packets being routed through the box), and OUTPUT (for locally-generated packets).

          nat:

          This table is consulted when a packet that creates a new connection is encountered. It consists of three built-ins: PREROUTING (for altering packets as soon as they come in), OUTPUT (for altering locally-generated packets before routing), and POSTROUTING (for altering packets as they are about to go out). Available since kernel 3.7.

          http://ipset.netfilter.org/ip6tables.man.html#lbAF

          My god, what's that "nat" table doing there then?

          1. Anonymous Coward
            Anonymous Coward

            Re: now just trollin

            I know it does but the IPv6 born agains first step is to wash away the sin of NAT.

        2. cjrcl

          Re: now just trollin

          NAT simply cannot scale indefinitely because of the 12 NAT layers limit inherent to BGP protocol specification.

          1. Anonymous Coward
            Anonymous Coward

            Re: now just trollin

            People can't scale indefinitely either (as long as chemical rockets are the best we can do). I guess the amount of shit we can sell them that data mines them does need to scale, so bring on the IPv6.

      2. Grifter
        Headmaster

        You can say irrespective or regardless, but you can't say irregardless.

    2. Number6

      Much to my surprise, Comcast rolled out IPv6. Somewhat better than VirginMedia in the UK, who seemed to be kicking the can down the road as much as it could.

      I did set the cable box into modem-only mode and have my own router (running OpenWRT) as my primary gateway. This was after Comcast did an 'upgrade' on the old box which broke Wifi and also reset it to default and I lost all my firewall settings. Now they can't. The IPv6 defaults to 'safe' because unless I make the effort to add a traffic rule, it won't pass IPv6 packets from the WAN to the LAN, so it's no worse than IPv4 and a NAT router with no port forwarding. That lets you use IPv6 outbound but still only be accessible on IPv4 if you've enabled any inbound stuff.

      The only irritating thing is that the IPv6 prefix changes occasionally, I guess it's the equivalent of re-segmenting the network and changing everyone's dynamic IPv4 address.

  5. DougS Silver badge

    Why is IPv6 faster?

    Is it faster for connection setup/teardown? Is it able to sustain faster throughput? Or are the higher layer protocols doing something different when on an IPv6 base to account for this?

    1. This post has been deleted by its author

    2. Dwarf Silver badge

      Re: Why is IPv6 faster?

      The choice of protocol doesn't not affect your bandwidth. Your local pipe will remain the same.

      Routing is significantly optimised in IPv6, that's what the first 48 bits are for. Its possible that some fat and shiny new pipes are only advertised over IPv6 (particularly in Asia)

      I expect the speed gain is a function of the protocol prioritisation in each of the OS's and this manifests in all the session set up and tear down. Consider how many connections a single web page makes as it pulls all the different components.

      I would not expect a significant gain in a single sustained TCP connection, but that's not what IPv6 is all about, its about fixing the broken things and work-around's in IPv4. Examples include the lack of address space, NAT, Carrier grade NAT, protocol optimisation and scaling, security (built in IPSec), which will in time help to reduce spam down as you'll know if you trust a connection long before any email can flow.

      Ask yourself this. Who still uses old connectivity such as RS485 / RS232, X21, X25, Arcnet, Token Ring, IPX, etc. etc. They are all largely dead technologies. The world moved on, embrace the change.

      1. Anonymous Coward
        Anonymous Coward

        Re: Why is IPv6 faster?

        > reduce spam down as you'll know if you trust a connection long before any email can flow.

        Wow you actually believe that don't you? Privacy get over it.

        1. Dwarf Silver badge

          Re: Why is IPv6 faster?

          @AC

          You mean that you believe that your public IPv4 address is "private" in some way today ?

          1. Anonymous Coward
            Anonymous Coward

            Re: Why is IPv6 faster?

            >your public IPv4 address is "private" in some way today ?

            Well I would like to think it is if resolves to a tor exit relay at least as far as my ISP is concerned. Granted its looking grim on that front but IPv6 certainly isn't going to enhance privacy in any way.

      2. Anonymous Coward
        Anonymous Coward

        Re: Why is IPv6 faster?

        Who still uses old connectivity such as RS485 / RS232

        Me… Hard to do Modbus/RTU or BACNet MS/TP over Ethernet and devices in the field that do Modbus/TCP or BACNet/IP are more expensive and can be more problematic.

        1. Anonymous Coward
          Anonymous Coward

          Re: Why is IPv6 faster?

          Also more than likely a significant amount of the semiconductors you used to post that message were created on tools still controlled over serial communication.

        2. Richard 12 Silver badge

          Re: Why is IPv6 faster?

          BACNet/IP is becoming more common than the older variety, most modern BMS-enabled kit I've encountered only does the IP version.

          (Which is a horrible protocol in fact)

          RS232 is still the number one multi-manufacturer industrial interconnect - because it's dirt cheap and trivial to secure. It can easily be unidirectional (cut one wire), and nobody is getting in unless they have physical access.

          Of course, managers then put an IP/RS232 bridge in and expose it to the Internet anyway.

          RS485 is the final stage physical layer for millions networks around the world. It's better than IP due to topology - multidrop is often far more useful than star.

          Often there is an IP to RS485 bridge, but the last mile (sometimes literally) is RS485.

        3. Paul Crawford Silver badge

          Re: RS485 / RS232

          Serial is still common in very low power devices as the overhead to implement a UART and the matching software stack is trivial in comparison. Seriously, work out the power used to be listening and able to respond quickly in both cases and you are in for a big surprise.

          Also simplicity makes for reliability/security as you are only point to point and not having to fend of the barbarian hoards when someone forgets to properly firewall your system.

      3. bombastic bob Silver badge

        Re: Why is IPv6 faster?

        "Who still uses old connectivity such as RS485 / RS232, X21, X25, Arcnet, Token Ring, IPX, etc. etc."

        RS232 and 485 are common with test equipment, as well as devices that have no ethernet. It's also common to have a serial port login for an embedded device so you can flash it with custom firmware (typical for GPL compliance), and of course things that have NO ethernet on them (like remote data collection devices). NOT being on the intarwebs or cellular networks is a form of security (so you'll see dial-up phone modems being used for data transfer with remote data collection devices).

        So: Even though you don't see a serial port on your latest PC, it doesn't mean they're not still being used... particularly by people working with microcontrollers.

        "The world moved on, embrace the change."

        Usually when I hear that, it sounds arrogant, like people who are Win-10-nic fans telling me I have to "embrace the change" or be pejoratively labeled a stone-age Neanderthal for NOT embracing it.

      4. Down not across Silver badge

        Re: Why is IPv6 faster?

        Ask yourself this. Who still uses old connectivity such as RS485 / RS232, X21, X25, Arcnet, Token Ring, IPX, etc. etc. They are all largely dead technologies. The world moved on, embrace the change.

        I do. So what? There is nothing wrong with using old tried and tested technologies. It's great world moves on, but change for change's sake is silly and not always practical.

        Disclaimer: This is a generic notion, not intended as argument on IPv4 vs IPv6

      5. Anonymous Coward
        Anonymous Coward

        Re: Why is IPv6 faster?

        "RS232, X21, X25, Arcnet, Token Ring, IPX, etc. etc. They are all largely dead technologies."

        Bollocks. E.g. there are still datacenters full of devices with RS232 serial consoles, and likely will be so for decades.

        "Old" doesn't mean "dead.

    3. ZeroSum

      Re: Why is IPv6 faster?

      IPv6 isn't faster, unless it has a lower RTT path than IPv4. As ISPs make their IPv4 & IPv6 peerings congruent and dual-stack CDNs the relative performance of IPv6 will continue to improve. Eventually ISPs and CDNs will put less effort into maintaining the IPv4 peerings and distribution of IPv4 enabled CDNs. Then the performance benefit that IPv4 enjoyed up until a few years ago will change in IPv6's favour. It is already happening from some operators to sites like Facebook.

    4. Anonymous Coward
      Anonymous Coward

      Re: Why is IPv6 faster?

      > Is it faster for connection setup/teardown? Is it able to sustain faster throughput? Or are the higher layer protocols doing something different when on an IPv6 base to account for this?

      The real reason is complex, but interesting - and for some reason doesn't seem to be discussed much in public.

      Try comparing the output of "traceroute" and "traceroute6" to the same destination, and you'll see that often the traffic takes completely different paths. Even if it doesn't in the forward direction, it may well in the reverse.

      Essentially, the IPv4 and IPv6 Internets are *two entirely separate networks*. Not only do they have their own protocols and numbering plans, they also have distinct topologies: some networks peer on IPv4 but not on IPv6, or vice versa. They really are ships that pass in the night.

      Now, for a long time one particular provider - one who also offers IPv6 tunnel services - has been trying to position themselves as a Tier 1 on IPv6 (a Tier 1 being a network which only peers with other networks, and does not pay for transit anywhere). To try to achieve this, they have been offering free IPv6 peering to pretty much every network at every exchange point, in order to collect lots of routes and make themselves an attractive peering partner to the other Tier 1's.

      The net result is that to reach certain networks your IPv6 traffic may go straight onto this provider's backbone from A to B, rather than going via A's transit provider and then B's transit provider via some intermediate peering point.

      This is clearly not a sustainable proposition - eventually this provider is going to have to have paying transit customers rather than giving free transit to both A and B. Also, to achieve Tier 1 it will have to get all the other Tier 1's to peer with it, which was proving difficult: at one point they even offered them cake (Google for "hurricane electric peering cake")

      So the upshot is: short term, IPv6 may be faster because some traffic may be using a different backbone for IPv6 than IPv4. Long term, this anomoly will vanish; IPv6 will then be slightly slower, because of the larger headers (40 bytes compared to 20 bytes)

  6. cjrcl

    About 6to4

    Home users as contrary to business users here in China have to employ 6to4 for IPv6 WWW. BTW, PPPoE over IPv6 capable routers is far from ubiquitous Worldwide.

    Have no ideal of why it is stated above that 6to4 together with Teredo constitutes no more than 0.01 % traffic. Although the Teredo portion admittedly should be omitted.

    1. ZeroSum

      Re: About 6to4

      Chinese and Taiwanese equipment vendors are supporting IPv6 network equipment and CPE that they are selling. e.g. Huawei and ZyXEL.

      Pure 6to4 needs a public IPv4 address so the host doing it can't be behind a NAPT.

      1. cjrcl

        Re: About 6to4

        Yes, I have a public IPv4 address but have no prerogative for an IPv6 subscription (for I am not a business user who is given more freedom on IPv6 as IPv4 Firewall is mature).

        Meanwhile, the router in use (Apple AirPort Extreme 802.11ac) which is a premium one is not PPPoE over IPv6 capable.

        Lots of people rely on 6to4.

        1. Anonymous Coward
          Anonymous Coward

          Re: About 6to4

          >for I am not a business user

          > the router in use (Apple AirPort Extreme 802.11ac

          You don't say?

        2. ZeroSum

          Re: About 6to4

          > Yes, I have a public IPv4 address

          Your ISP is going to run out of public IPv4 addresses. They'll have to use so-called Carrier Grade NAT.

          The best ways of doing that use IPv6 as the underlying transport. Eventually you'll get IPv6.

          1. Anonymous Coward
            Anonymous Coward

            Re: About 6to4

            Even if someday you are forced by your ISP to use an IPv6 address on your router it doesn't mean you have to go and enable IPv6 on everything you own.

            1. cjrcl

              Re: About 6to4

              If I would be forced to IPv6 by ISP. An upgrade of my router should be a must unless something anther than PPPoE be deployed.

              1. Anonymous Coward
                Anonymous Coward

                Re: About 6to4

                I would say PPPoE is the hack that indirectly built the cable industry into what it is today but that is actually the silly French insisting ATM use 53 byte padded cells (just big enough not to fit a Ethernet TCP/IP header). Brilliant.

                1. Adam 52 Silver badge

                  Re: About 6to4

                  "ATM use 53 byte padded cells (just big enough not to fit a Ethernet TCP/IP header"

                  What!?

                  You are aware that ATM is quite capable of operating without IP aren't you? Only this morning I was on a dodgy video conference wishing for a decent, real-time protocol that doesn't require shed loads of over overprovisioning.

                  1. Anonymous Coward
                    Anonymous Coward

                    Re: About 6to4

                    >You are aware that ATM is quite capable of operating without IP aren't you?

                    Quite yes and its actually pretty good for real time as you say but good luck trying to get new (but oh so old) POTS service any more in many places. You will take VOIP and like it because that is all there is.

                    1. Vic

                      Re: About 6to4

                      good luck trying to get new (but oh so old) POTS service any more in many places. You will take VOIP and like it because that is all there is.

                      And blaming ATM for that makes as much sense as blaming cheese...

                      Vic.

              2. ZeroSum

                Re: About 6to4

                >An upgrade of my router should be a must

                ISPs are not making any money off the transition to IPv6.

                That is why they have been so slow to adopt it.

                When you get new equipment from your ISP it will be because you changed provider, bought a new service, or the modem broke. IPv6 is a technical nice to have as far as the business case for deployment goes.

            2. ZeroSum

              Re: About 6to4

              > Even if someday you are forced by your ISP to use an IPv6 address on your router it doesn't mean you have to go and enable IPv6 on everything you own.

              As IPv6 becomes more established any problems that make people want to turn if off will be ironed out.

              As Metcalfe's law kicks in for IPv6 there will be an increasing incentive to use it.

          2. cjrcl

            Re: About 6to4

            OMG, Carrier Grade NAT, then I have to move to my company to get rid of the State Firewall :(

        3. The Original Steve

          Re: About 6to4

          "Meanwhile, the router in use (Apple AirPort Extreme 802.11ac) which is a premium one ..."

          You're confusing "premium" with "expensive".

  7. Nate Amsden Silver badge

    20% is not noticable

    20% of anything generally is a rounding error for me. 20% faster, 20% cheaper, 20% more efficient.

    I just checked my AT&T phone again, it is not roaming, and is assigned an IPv4-only IP 10.xx address. www.whatismyipaddress.com says I am coming through a proxy server(or perhaps carrier grade NAT maybe they don't differentiate). Never had a problem accessing things even with tethering.

    I just spent 3 months in thailand, didn't happen to notice what my phone with a local SIM card was assigned if it was v4 or v6, but more than anything the latency(purely distance I imagine) killed performance, I managed to get somewhat better web performance tethering and going through a VPN and proxying my HTTP traffic through my co-located server in California, though not for geo-diverse sites).

    Myself I have no interest in IPv6, I do not have any plans to deploy it on any of my networks, if it happens to get deployed in a transparent fashion upstream on my mobile or broadband connection I don't care since it is transparent (I do not make inbound connections to my phone, and any inbound connections to my broadband are done through a VPN established outbound to my co-located server so I don't need static IP and don't have to care about NAT that the ISP might be doing).

    I figure if at some point I really need to support inbound IPv6 to the websites that I manage then I'll have that translation handled by the CDN (just like they already do things like SSL termination and TCP optimizations already).

    The die hard IPv6 fans remind me of the modern web developers who want you to jump on the new thing just because.

    I would put IPv6 up there along with Software defined networking(SDN) as something that can benefit large scale companies well(mainly service providers), but for 95% of the downstream organizations it provides no benefit.

    IPv4 addresses are available, getting large blocks of them may be difficult, though every company I have worked for in the past decade hasn't needed more than a few dozen external IPs at most (even today not hard to get). Smart use of NAT and name-based virtual hosting(with good proxy servers - in my case Citrix Netscaler or F5 BigIP) goes a long way.

    Now that SNI is pretty widely adopted that removes a large reason to need a big number of IP addresses on servers with multiple SSL certs(covering multiple domains so a wildcard cert wouldn't be sufficient).

    Die hard IPv6 proponents say IPv4 is full of hacks and brokeness so we should upgrade because of that. The hacks seem to work fine, I suppose if you are directly suffering from those limitations then you should probably upgrade, myself, I really haven't had any pains associated with IPv4. For example the last time I was at a company that needed a site to site VPN with another company and we happened to have overlapping IPs so had to do 1:1 NAT -- that was literally 11 years ago).

    Myself I don't do peer to peer anything and the VoIP solutions I have used seem to work fine with IPv4 and NAT etc.

    So no pressing need to upgrade, I don't see a pressing need for many years to come(for folks like myself anyway).

    1. This post has been deleted by its author

    2. DuncanL

      Re: 20% is not noticable

      20% of anything generally is a rounding error for me. 20% faster, 20% cheaper, 20% more efficient.

      Can I have 20% of your pay cheque please?...

      1. Anonymous Coward
        Anonymous Coward

        Re: 20% is not noticable

        Though I agree with what he says I was going to go with the kicked 20% harder in the crotch analogy but what you said is a bit more eloquent.

        1. Nate Amsden Silver badge

          Re: 20% is not noticable

          I just moved away from the bay area for a 50% drop in housing cost which was approaching 50% of take home pay.(up over 1000/mo in past 5 years).

          You joke like 20% is a lot, it's all relative of course. I routinely tip 30 or 40% for good service. If it's really good maybe I go 60 or 75%. On a few occasions go over 100%.

          My mother asked why I wasn't pissed off that the government takes roughly 45% of my pay. I told her that's just how it is. Doesn't really bother me since it's pretty consistent (unlike housing costs).

          My new apartment has internet speeds that are 10 times faster than my previous apartment for about the same cost. I will notice that for sure.

          1. TheTick

            Re: 20% is not noticable

            Do some work with local government or some government agency and see how they spend that money they take from you.

            You might not have the same opinion afterwards (I didn't).

    3. SImon Hobson Silver badge

      Re: 20% is not noticable

      Downvoted for this :

      I don't see a pressing need

      That's the point, you don't SEE a neede, that doesn't mean it's not there.

      What you don't see is all the wasted work, wasted cost, etc involved in making many things work with IPv4 and NAT. You fire up a program is "it works", you don't see the effort the dev had to put in to make it work.

      As someone who's dealt with installing and troubleshooting various stuff for many years, I've seen the problems NAT causes, and the hacks needed to work around it - or in some cases, not work around it (thanks Zyxel and your crap-by-design NAT). Not even FTP works through NAT without help - but people don't see that because pretty well every NAT gateway also includes an FTP helper to fiddle a few bits and work around the breakage.

      Noticed how almost all these "control your ${something} from your phone" systems all use a hosted server ? It's only partly so the vendor can slurp your data and sell you to advertisers - it's also to work around the problems caused by NAT (and to an extent, dynamic IPs).

      VoIP - yup, problems with NAT.

      Peer-peer stuff like torrents - yup, problems with NAT.

      Yes, these are all surmountable problems (for example, the voip provider we use at work has large proxy servers - the cost of which goes into the product) - but the point is that most people don't generally see them because other people have wasted effort in working around the problems. All this effort is largely a waste when IPv6 simply removes those problems.

      So anytime I read a tirade like yours about how "it's not needed", I know it's someone who hasn't got a clue and really has no idea how much borkage NAT causes and the effort needed to work around it.

      1. Vic

        Re: 20% is not noticable

        You fire up a program is "it works", you don't see the effort the dev had to put in to make it work.

        Unless you're trying to open a server to the Internet, that effort is zero to negligible. And if you *are* trying to open up that server, I *want* there to be some effort to prevent services being accidentally exposed to the outside world.

        Not even FTP works through NAT without help

        PORT mode doesn't work through NAT - and that's a good thing. PORT mode is obsolete. PASV mode works just fine.

        Noticed how almost all these "control your ${something} from your phone" systems all use a hosted server ? It's only partly so the vendor can slurp your data and sell you to advertisers - it's also to work around the problems caused by NAT (and to an extent, dynamic IPs)

        It's a zero-config way of setting up such kit; the manufacturers have cottoned on to the fact that end-users don't like having to configure everything. Once (if) we get rid of NAT, things will be no different - end-users won't like having to punch holes in their firewalls, so manufacturers will carry on doing exactly the same thing. This is not a NAT/IPv4 issue, it's an end-user one.

        VoIP - yup, problems with NAT.

        I wish people would stop quoting that as if it were gospel - I've been running (and reselling) VoIP services through NAT for over a decade. It works trivially. The only time you ever come across problems is when certain manufacturers (Juniper, most especially) try to add "helpers" to their networking kit to overcome some perceived issue; turning all that shite off makes it work.

        Peer-peer stuff like torrents - yup, problems with NAT.

        Ten years ago, perhaps. These days, it just works.

        Yes, these are all surmountable problems

        For the most part, they're non-existent problems.

        (for example, the voip provider we use at work has large proxy servers - the cost of which goes into the product)

        That's their choice of architecture; it's not the only one. Once the session has been initiated, individual phones can STUN their way through a NAT router and communicate directly - but that makes billing much harder, which is why commercial VoIP providers use a proxy. If you're not billing the calls, you don't need the proxy...

        All this effort is largely a waste when IPv6 simply removes those problems.

        Nope. IPv6 changes the problem space. End-users will have to manage their own firewalls. Anyone who thinks that is a good idea has never done home support. Overall, geeks will be better off with IPv6, less-technical users will find IPv4 much easier.

        So anytime I read a tirade like yours about how "it's not needed", I know it's someone who hasn't got a clue and really has no idea how much borkage NAT causes and the effort needed to work around it.

        Well, there are definitely two sides to that story. You are also entirely ignoring a set of very real problems that others will face because you don't see them as significant. IPv4 has some significant problems, but claims of rainbows and unicorns aplenty when we move to IPv6 are blinkered; IPv6 has its own problems, they're just different to the IPv4 ones.

        Vic.

        1. Neil Alexander

          Re: 20% is not noticable

          "And if you *are* trying to open up that server, I *want* there to be some effort to prevent services being accidentally exposed to the outside world."

          That is the job of a firewall. Repeat after me: NAT is not a firewall.

          1. Vic

            Re: 20% is not noticable

            That is the job of a firewall

            Yes. Now try to get your average home user to understand how to program a firewall. What will happen is that it gets switched off, because that "makes it work".

            Repeat after me: NAT is not a firewall.

            However often you repeat your mantra, it affords a deal of protection for those that do not understand networking. Taking that protection away - even if there are better systems - does not improve anyone's life; to use those better systems requires an amount of understanding that most computer users just do not have.

            Vic.

  8. John Sanders
    Facepalm

    I'm sure many of the punters here

    Have never had to set-up IPV6 infrastructure from scratch, or have to administer any large IPv6 network.

    It is an abomination.

    1. ZeroSum

      Re: I'm sure many of the punters here

      > Have never had to set-up IPV6 infrastructure from scratch, or have to administer any large IPv6 network.

      I have hundreds of routers with IPv6 and it was easy.

    2. bombastic bob Silver badge

      Re: I'm sure many of the punters here

      (re: set up IPv6 from scratch)

      "It is an abomination."

      well, it's DIFFERENT that's for sure. once you understand, not so hard. just make sure you firewall it properly. Someone else besides me already mentioned that last part...

      One problem I ran into was an older WiFi router that aggressively kept trying to route things via its own WAN port, even if I shut OFF the IPv6 support on it. My solution: assign static IPv6 to both the LAN and WAN interfaces, then use an ethernet cable to plug the WAN into the 5 port LAN side. That way, if it insists on routing, it's back onto the network. [biggest problem was that it affected things on the ETHERNET side too, not just the wifi].

      So you have PROBLEMATIC DEVICES like *that* as well, announcing themselves as 'routers' whether you want them to or not, then B0RKING the routes that aren't statically configured, because they're so aggressive and effectively unconfigurable.

      In any case, as long as you're using Linux or BSD as your gateway, you should be able to set the daemons up that you need to make it all work, with plenty of instructions on how to do so online.

    3. SImon Hobson Silver badge

      Re: I'm sure many of the punters here

      It is an abomination.

      And so is IPv4. I assume you've also never setup IPv4 stuff from scratch either - and I don't mean the "from scratch" available today with largely auto-provisioned routers, LAN settings already there and DHCP enabled, etc. Having done it when manual config was still common, SLIP lines were the norm, etc - IPv4 was no piece of cake back then either.

      But, IPv6 really isn't any harder - it's just that people are "comfortable" with IPv4 and so it seems harder. The biggest obstacle is getting away from the "it's the same with longer addresses" mindset since some things are different - not necessarily better or worse, but different. Part of this is also getting away from the "networks start and end with Ethernet" thinking some people have because they've never dealt with anything else.

      E.g. The concept of "on-link" neighbour is very different - but once you get to grips with what is different and WHY then actually I do think it's much better. But if you can't escape from what you know about IPv4 then you will find it hard - but that's you making it hard when it isn't.

      PS - I've been IPv6 enabled both at work and home for several years (via a HE tunnels) , and pretty well "it just works" and I seldom notice it.

  9. The Average Joe

    Cisco and Charter internet have it done

    It is NOT IPv6, it is double NATed IPv4 and then carrier grade nat as it exits the charter network in Detroit. Each static subnet cable modem has 2 paths, one paths is the static address pool and the other goes out the double NATed carrier network. Prolly using Cisco FabricPath

    If Cisco was smart they would come out with IPv7 that fixes all the complexity of IPv6 but allows the use of IPv4. Dual stack OSes suck, remember Windows NT - IP and Netware - NCP?

  10. Kevin McMurtrie Silver badge

    Actually, thank the big telcos

    As much as I hate US big telcos, they are the reason that IPv6 works. I have yet to see a decent home or small business router that can work properly with IPv6 turned on. You fight your way through inconsistent terminology, auto-configure bugs for the subnet size, and numerous things that need a reboot for no good reason. You start wondering if "6rd" is pronounced as "turd." Then connections start timing out because somewhere in the 1000 miles between point A and B, the MTU is smaller than each side thinks it is. Auto-config puts your MAC address in the IP address, then another auto-config tool warns you that your MAC address is visible. Maybe router rules doesn't work for IPv6 so you have to disable IPv6 on everything that's not hardened against global access.

    Plug in the big telco modem and IPv6 usually works.

    1. ZeroSum

      Re: Actually, thank the big telcos

      Agreed that the big US ISPs have helped drive an improvement in CPE but that much of the CPE is still very poor.

      > Auto-config puts your MAC address in the IP address, then another auto-config tool warns you that your MAC address is visible.

      It is the host not the router that forms the host bits of the IPv6 address with SLAAC. Modern OSs no longer by default use EUI-64 addresses for communication beyond the router. They use privacy addresses.

  11. Christopher Slater-Walker

    Not entirely on topic, but...

    I'm i

    on BT Infinity at home, but using a Cisco router rather than the HomeHub. I recently discovered that I've been IPv6 enabled, and configured the router accordingly to get the /56 delegated prefix. All well and good, or so I thought...

    After a test reboot of the router, I find I get a different /56 prefix. It's not static.

    I raise a complaint with BT, who tell me that there is no way they can do static addresses on domestic broadband. They can't even tell me if the IPv6 prefix will be static on a business-type connection. I'm waiting for them to find out. (The IPv4 address would definitely be static).

    Since we're all supposed to be moving to a world with countless connected devices at home, at work and indeed everywhere. How will I be able to access my house controls etc. on IPv6 when I don't know what their address will be from one day to the next? I find this decision by BT to be incredibly stupid and short-sighted.

    1. ZeroSum

      Re: Not entirely on topic, but...

      > They can't even tell me if the IPv6 prefix will be static on a business-type connection. I'm waiting for them to find out. (The IPv4 address would definitely be static).

      Of course they could do static IPv6 delegated prefixes, they just don't want to do it yet.

      It will be easier to implement static IPv6 than static IPv4 because they'll have mountains for virgin IPv6 address space to assign from.

    2. SImon Hobson Silver badge

      Re: Not entirely on topic, but...

      who tell me that there is no way they can do static addresses on domestic broadband

      Of course there is, what they would say if they weren't a bunch of ****ing ****ing ****s is that they won't do because they want to force you to pay extra (double). They figure that if you want a static address then you'll pay for it - even their business broadband has dynamic addresses unless you pay extra (last time I heard it was £5/month extra !).

      So no there is no "can't" about this, it is all about "we won't because we want to screw you over".

      Same with Sly and StalkStalk - they won't do fixed addresses on "consumer" broadband either.

    3. Vic

      Re: Not entirely on topic, but...

      I raise a complaint with BT, who tell me that there is no way they can do static addresses on domestic broadband. They can't even tell me if the IPv6 prefix will be static on a business-type connection

      That's not a problem with IPv6, that's a problem with BT. They can - and should - give you a static prefix (it's not like there's any scarcity of those). But they have chosen to be a bunch of clueless numpties, and there's no fixing that with technology...

      Vic.

  12. geeky_poor_musician

    Netflix...

    It's ironic that Netflix is noticeably faster on IPv6, since Netflix has been blocking my IPv6 connection attempts for the last two months or so.

    Had to firewall off Netflix's IPv6 pool (2a01:578:3::/64) in order to get around the Geo fence.

    Yes, I'm in the UK, and yes, I have a UK subscription.

    How irritating!

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Biting the hand that feeds IT © 1998–2019