Registries differences

[John Curran]

At 17:02 4/29/97, Jim Fleming wrote:
>...
>I will continue to try to fight for the small
>companies that are just getting started
>in the Internet. In my opinion, those
>people are the future of the Internet.
>Any model developed for small companies
>could easily be used with large companies.
>I am not sure the reverse is the case.

Full agreement.  One of the reasons that I think 
a provider-governed organization such as ARIN is
so important for IP address allocation is so that
we do not end up with any number of alternatives
which only favor large companies.  

If we as an industry do not provide for responsible
management of the IP address space, then it is quite
possible that we will end up with a far more colorful
goverment/regulatory solution in this area.  I'd love
to think that such a result would be equitable and fair
to all players, but common sense dictates that such 
outcomes are generally deaf to the cries of smaller 
companies and pliable to the firms with far greater 
resources and access.

>While I might agree that the BBNs and
>NSIs of the world could now take the
>Internet and divide it up along the lines
>that you describe, I will try to make the
>point that small companies should be
>given a chance. I understand why you
>do not want that to happen.
>...
>ARIN as an extension of NSI/SAIC and
>backed by BBN and other large companies
>does not fit the model that I think best
>serves the Internet community.

I'm uncertain that ARIN backed by "large companies"
as much as "folks at large companies who have an
interest in seeing the Internet continue to grow". 
It's quite likely that the large backbone providers
would come up with a very different proposal than 
ARIN if given a chance, but as luck would have it,
many the people who would be involved in preparing 
such have been part of the cooperative Internet for
a long enough period of time to know that simple &
open solutions encourage growth. ARIN is attempt to 
create such a system for IP block allocations, and  
while it may not be perfect, I have yet to see an 
alternative which comes close.

I am certain that we can create far more complex,
costly, and arcane processes for performing this 
relatively simple task.  If we fail as an industry
to manage this problem, then we will get to explore 
many of the more interesting alternatives with 
ample outside guidance.

/John