Pengshuping (Peng Shuping) wrote:
I just wonder, over two decades ago when the discussions happened,
whether the question that was asked most was " Why do we need to go
with 128 bits address space if whatever is been trying to achieve
with the existing approach of IPv6, can be achieved by 64 32bits
(IPv4) address space as well?"
The discussion, then, was whether we should use fixed 64 bit address
or something else and we mostly agreed on fixed 64bit.
Then, all of a sudden, IPng directorates determined that address
should be 128bit long without much reasoning.
Though some directorates later mentions all the wrong reasons to
do so, which means they do not have proper expertise to make
such decisions, the true reason was to make lower 48bit of IP
address MAC address, which is a very old idea deployed by XNS,
which is mostly abandoned even in IPv4 era obviously because
of unnecessarily lengthy and irregular addresses.
Maybe, some thought, unlike XNS era, DNS could help without
muchknowledge on the reality of network operations.
Just a bit curious about, why NOW when today’s technologies (HW/SW
processing capabilities as well as the ever-increasing bandwidth) are
more advanced compared with those at 25 years ago, suddenly people
become very concern with the overhead and start questioning about the
“extra burden” caused by the packet address/header length…
Not suddenly. 128 bit address has been recognized as major obstacle
to deploy IPv6, because remembering 8 chunks of 4 hexadecimals is
divine, where remembering IPv4 addresses is human.
Masataka Ohta
