Announcement

Collapse
No announcement yet.

IPv10 Draft Specification Published

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #51
    Originally posted by clockley1 View Post
    We do not maintain web servers.
    Who's this "we" you're talking about?

    Comment


    • #52
      Originally posted by duby229 View Post

      It's a process called port forwarding. I think most modern routers though call it something else and do it automatically.
      Port forwarding is crap, which gets further complicated as a number of such users, agents and aplications increases.

      Also, port forwarding demands router setups on both sides, which is not always available.

      With IP6, these and similar things are a breeze. Static IP is never an issue. You usually get not one, but a few quintillions ( they have gone way overboard with this one IMO ). It's not that hard to make sections on firewall so user gets adresses for local communication, another spread for global visibility, yet another for connections with ones peers etc.

      Last but not least, PF demands that router mangles packets- manipulates with IPs and ports, checksums etc. This can cause bottlenecks.
      With IPv6, it just has to forward packets.


      Comment


      • #53
        Originally posted by Brane215 View Post
        You usually get not one, but a few quintillions ( they have gone way overboard with this one IMO ).
        IIRC the idea originally wasn't that anyone would have a need to have a zillion hosts in a subnet, but rather that the host part of the ipv6 address would just be the MAC address of the interface. Thus getting rid of the requirement to manually allocate IP addresses, or run a service such as DHCP.

        Of course, later on people realized that having your MAC address be part of your public IP could be a privacy issue, and various protocols were invented which allowed hosts to generate a random host part of the address or something like that. But also in this case the large address space is useful, since it's exceedingly unlikely you'll run into a collision.

        But anyway, yes, NAT and port forwarding needs to die in a fire. The sooner the better.

        Comment


        • #54
          Originally posted by duby229 View Post
          It's a process called port forwarding. I think most modern routers though call it something else and do it automatically.
          Won't work on most mobile networks, and on many ground networks as there are double or even triple NATs and you have no control over your ISP's 2 or three NATs upstream.

          There is a reason if services like Skype and Teamviewer act like total haxxors (do weird stuff with ports and if all fails they pipe their connection over port 80 or 433, http/https ports and then are routed to their own servers outside of whatever random shit your ISP has done).

          Comment


          • #55
            Originally posted by torsionbar28 View Post
            Who's this "we" you're talking about?
            that part is quoted from debianxfce, and the guy asked proof of performance increase if you disable ipv6.

            I'm 100% with you. I have my own private cloud so yeah I run my own mini server and had to pay more for a public IP (ipv4 as the only ISP I can use is retarded apparently) because I'm behind n levels of NAT on the ISP side and no goddamn thing works with port forwarding.

            Comment


            • #56
              Originally posted by DMJC View Post
              The IPv6 standard is dumb. They should have just added extra octets to v4. 192.168.xxx.xxx.xxx.xxx.xxx.xxx.0 makes a hell of a lot more sense than retraining everyone on yet another dumb standard.
              No, it's better to have them very different. Not like ipv4 netmasks (and subnetting) ever made any sense.

              Comment


              • #57
                Originally posted by debianxfce View Post

                We do not maintain web servers. To make your kernel cleaner and faster, remove ipv6 support.
                You are a moron. To make Internet cleaner and faster, remove "debianxfce" troll support

                Comment


                • #58
                  Originally posted by bug77 View Post
                  They are when you factor in customer support
                  There are like 3 different ipv4-in-ipv6 technologies and they all work fine (for the average user needs anyway). Most consumers won't notice, because they don't need to do anything on their side.
                  Newer deployments get a router with IPv4 LAN and a IPv6 WAN port.

                  Comment


                  • #59
                    Originally posted by starshipeleven View Post
                    No, it's better to have them very different. Not like ipv4 netmasks (and subnetting) ever made any sense.
                    Also, it's not like they were working on just one problem - way bigger address space, but they were also put significant amount of work into addressing other issues with IPv4.
                    One of them was for example load that IPv4 presents on intermediate nodes etc.

                    IPv6 is definitely here to stay ( if nothing else, there is simply no alternative) and I can definitely see more and more users migrate every day.


                    Comment


                    • #60
                      Originally posted by starshipeleven View Post
                      There are like 3 different ipv4-in-ipv6 technologies and they all work fine (for the average user needs anyway). Most consumers won't notice, because they don't need to do anything on their side.
                      Newer deployments get a router with IPv4 LAN and a IPv6 WAN port.
                      Why would you need Ipv4 LAN these days? Just because you want to browse from MS-DOS?

                      Comment

                      Working...
                      X