Encouraging return of legacy space WAS Re: [ppml] ARIN Policy Proposal 2002-9

[Mury]

> While it may be a good idea to contact holders of large amounts of legacy 
> IPv4 addresses in the Class A and Class B ranges to see if they will 
> voluntarily return the space to IANA, this type of thing should not 
> concern ARIN at this time. This is the sort of activity that should be 
> coordinated globally and it should be done with the blessing of IANA and 
> ICANN if it is done at all.

Hasn't IANA and ICANN basically given ARIN the task of managing IP space?  
Perhaps I have missed something.  Of course it should have IANA's and
ICANN's blessing.  Why should ARIN have to be dictated first by ICANN?  I
have no clue what goes on at ICANN, but I have a suspicion that the people
in and around ARIN are more in touch with these issues.

Let me put it to you this way... A father asks that his teenage boy to
help with more of the yardwork.  "Where do I start dad?"  The father
instucts the boy how to mow the yard.  Upon mowing the yard the boy sees
that by trimming some of the weeds around the fence the condition of the
yard would look much better.  The boy decides to ask his dad if he can
trim some of the weeds that the mower couldn't reach.  Dad is pleased.  
The boy is pleased.  The neighbors are pleased.

Obviously the weeds are those unused legacy IP blocks.  ARIN should draw
up a simple game plan, get all the appropriate parties blessings, and at
least make another attempt to get that IP space back.

If a company needs a smaller block, I say give it to them for free.  If in
the next 5 years they find out they need more IP space, give that to them
for free too, but have some sort of time limit on the free IP space.

As far as your global comment, I doubt that many African, Asian,
Australian, South American, etc. entities are sitting on lots of large
unused blocks, but I certainly could be wrong.  And what if I am?  Why
can't ARIN still do it?  Why can't they become a model for the process if
it works?

> ARIN members would be better off deploying testbed IPv6 networks and 
> connecting these with other company's testbed IPv6 networks in order to 
> gain experience with IPv6. When the address shortage does become 
> significant enough to push customers towards IPv6, it is likely to quickly 
> become a tidal wave migration. Historically, IPv4 was deployed for many 
> years before there was a sudden exponential demand for it in 1994 and 
> 1995. The main reason that ISPs were able to meet the demand is that there 
> were a lot of people around who had years of hands-on experience running 
> IPv4 networks. At the time they were gaining that experience, they didn't 
> know how IPv4 would be used in the future. They didn't foresee the web and 
> e-commerce and then decide to run an IPv4 network. Similarily, with IPv6 
> we don't need to predict grand future horizons in order to find a reason 
> to deploy it today. One small reason is enough and it doesn't have to be 
> the same reason for everybody.

Are you saying that ARIN's resources are so strapped they can't blink and
breath at the same time?  I find this extremely hard to believe.  If
testing IPv6 in the manner you suggest is agreed to be useful, why can't
ARIN do both?

Regards,

Mury