On Thu, Feb 16, 2017 at 5:40 PM, <otroan(_at_)employees(_dot_)org> wrote:
See 7421, 6177, 7368, section 3.4.
There are many reasons for the 64 bit boundary.
- Allowing identifier locator split: 8+8 / GSE that led to ILNP and NPT66
- Simplicity in addressing (no more subnet masks)
- A fair balance between the users and the providers of networks.
Ensure that users get a fair share of addresses and try to avoid
operators charging per address.
The 64 bit boundary is so embedded in the set of IPv6 specifications that
it would be very hard to unravel at this point. It certainly cannot be a
single paragraph put in during the advancement of 4291.
Right. If we want to change this we can, but the process for that is start
with "write a draft and get consensus in 6man". We cannot change it while
revving 4291 to full standard, because it's a substantial change.