-4

128 bits is much less user readable than a six-byte dotted format.

Retaining a dotted format could have led to less drastic changes to both low level and application software.

Talljoe
  • 14,593
  • 4
  • 43
  • 39
Fubarro
  • 1
  • 1
  • 1
    Readability is not a valid argument here. – Jeremy Visser May 06 '11 at 05:37
  • Or rather, use DNS that's what it is for, stop hard coding IP addresses. – Steve-o May 06 '11 at 14:49
  • 128 bits was a sensible choice given the information available at the time. But considering some of the things we have learned since then, it should have been longer. Why does Teredo only support a single host on a tunnel and not a full network like 6to4? Because it would have needed 200 bits to do so. Proper use of RFC 4193 only gives you 16 bits for subnetting, but you can squeeze 21 bits out of RFC 1918. Some systems make use of embedding a cryptographic hash inside an IPv6 address. But 128 bits doesn't make for a very strong cryptographic hash. – kasperd Jun 21 '15 at 21:59

3 Answers3

6

IPv4 to IPv6 is a backwards-incompatible change no matter how you slice it, so there's no reason to do a half-assed job of it. Having a REALLY wide address has all sorts of interesting benefits anyway, like simplifying routing and address autoconfiguration.

1

Future-proofing. The v6 transition is a big enough headache that no one wants to ever have to switch to IPv7. The dotted format is mainly for the benefit of humans anyway, not the machines.

Jumbogram
  • 2,249
  • 1
  • 20
  • 24
0

We wouldn't have had enough time to fight off the nanobots that way.