RE: Fixed IP / IP ranges

[Kevin Meynell <kevin@xxxxxxx>]

>In theory you can have over 4 billion addresses which will see us fine
for
>a few years yet.

The problem is not so much the number of IPv4 addresses available, but how
they've been distributed. For example, there are shortages in amount of
address allocated to the Asia-Pacific region (administered by APNIC) when
you consider there are potentially 4 billion users in that region alone.
However, the immediate problem is routing.

The IPv4 address blocks are quite fragmented which means the Internet
routing tables have to be larger than necessary. This was a serious problem
before the introduction of CIDR a few years ago (when the Internet came
very close to meltdown), but the problem is not getting any better.

Of course, some of the unused blocks could be reclaimed and the address
space reorganised. Unfortunately, that would take almost as much effort as
the transition to IPv6, and without offering the advantages of IPv6.

>The only question in my mind is whether this is incompetence or
deliberate
>waste in an effort to force IPv6 before it's really necessary.

Errr.. well given that the most wasteful allocation of IPv4 addresses took
place before IPv6 was even invented, my vote is for incompetence! It was to
be remembered that no-one envisaged the phenomenal success of the Internet,
and you could equally apply the same question as to why we need to have
11-digit telephone numbers in the UK, when 8 would be perfectly sufficient.
The answer is because the original numbering plan was devised around the
technology available at the time, and would take too much effort to start
again from scratch.

>"It's being deliberately created by groups with financial
interests in a
>move to IPv6.

Well of course it is - they're not doing it for love! However, the unwanted
side-effect is that it solves a number of problems with IPv4 ;-)

Cheers,

Kevin Meynell