On Oct 15, 2012, at 14:15:22, Brian Wood <woodbrian77 at gmail.com> wrote: > Erik Anderson: > > > Absolutely not. > > > > Making a change like this is a *big deal*, both in terms of money (to > > upgrade/replace network infrastructure) as well as in terms of having > > to learn a new technology. As such, it was very wise for the IP > > governing boards to not just make an incremental bump in the IP > > address space, but make a *huge* increase. This decision ensures that > > we won't need to go through this whole process again in the > > foreseeable future. > > > 8 bytes is a huge increase. IPv4 has lasted longer than expected > so I can't imagine 8 bytes being exhausted in the future. I read that > 16 byte addresses can address more atoms than are thought to exist. > From a practical point of view I think the 16 byte addresses are > a mistake. Systems have to work through all of that before they > can start to do something useful. That's a good reason not to > switch to IPv6. I believe you about IPv6 being an improvement > over IPv4 in a number of ways, but think the length of the addresses > was a mistake. > > I don't think anyone is paying for IPv6 specific upgrades to hardware. > When they upgrade for a practical reason, the hardware they get is > more IPv6 capable than what they had. Part of IPv6 is to use the existing MAC address for auto-configured addresses. Being that there are already supposed to be enough of THOSE for everything to be uniquely addressed, then you have to add network addressing on top of that, it was probably just 'easier' to increase the address space to accommodate those. ----- Eric F Crist