In elaborate terms: you have the ability to change any one of the protocols, specifications, designs or standards of the above at their proposal stage or before their mass adoption. You may choose to modify or reject an existing one or create one by yourself.

Some users and I would have common ideas in mind, however I would love to see some esoteric ideas as well.

  • Boomer Humor Doomergod@lemmy.world
    link
    fedilink
    English
    arrow-up
    18
    arrow-down
    3
    ·
    1 day ago

    Stop IPv6 from existing.

    Make IPv5, add a fifth number to the address, and improve NAT.

    Not every particle in the universe needs a publicly routable address.

    • ambitiousslab@feddit.uk
      link
      fedilink
      English
      arrow-up
      10
      ·
      1 day ago

      That’s interesting - I hadn’t heard too much dissatisfaction with IPv6 before, except for the slow adoption, and the not-as-nice looking addresses. Is it an aesthetic preference or just that IPv6 is overkill? Or any other advantages to doing it the “IPv5” way?

    • palordrolap@fedia.io
      link
      fedilink
      arrow-up
      5
      ·
      24 hours ago

      IPv5 existed. It was called the Internet Stream Protocol. The fact IPv4 used 4 octets was a happy coincidence more than anything, so v5 wouldn’t necessarily imply a ninth chevron fifth octet.

      But IPv4+, whatever that might have been, could have been an extensible system like, say, Unicode, and taken advantage of the unallocated/reserved 240.0.0.0/4 block to flag that the address is longer and the rest is encoded elsewhere in the packet.

      I mean, if you want to go completely crazy, you could specify ~2^28 further octets with such a system… although requiring a 256+ megabyte MTU might be slightly too extreme.

        • palordrolap@fedia.io
          link
          fedilink
          arrow-up
          1
          ·
          45 minutes ago

          They weren’t thinking big enough. They’ve only doubled the address space. I say this at least half seriously, well aware that mine is far more ridiculous the other way.

          … but I probably should have tried searching for “IPv4+” before using it as a generic term. At least one other proposal shows up when I search for that, and one of them is a proposal that adds a couple more octets.