[ppml] POC verification process

[Ian Baker]

----- Original Message ----- 
From: <Michael.Dillon at radianz.com>
To: <ppml at arin.net>
Sent: Tuesday, August 12, 2003 12:19 PM
Subject: Re: [ppml] POC verification process


> >Just sticking to the 10 range in IPv4.. 16 million large organizations is
> >not enough?
>
> The problem is that it is unregistered and there is no way to control what
> ranges any organization might be using internally. If it were really
> feasible
> to build an internet without globally unique IP addresses, then ARIN would
> not even exist.

Well, it seems to me that you are proposing an independant proprietary
network addressing scheme.

Since this is, by your own definition (it would appear to me) independant of
the Internet, then I can't see the problem in simply using the 10 block and
maintaining an independant record, one IP per defined organization. The
organization could presumably just NAT from their network into yours.

> >I dare say that IPv6 would solve the problem entirely.
>
> Yes. Because of the large address space, it is straightforward to get a
> large
> enough allocation to build a sizable internetwork. But v6 isn't here yet
> because
> the application developpers aren't pushing or supporting it.

There's a reason for that. Well, several really.

> >after all, you're
> >talking about a strictly private and proprietary network,
>
> No. An internetwork, by its very nature, is not strictly private. There
> are
> many private companies who are sharing the internetwork just like the
> public Internet. The main difference is that the AUP of the public
> Internet
> is very loose and only prohibits network abusers from connecting. On the
> financial industry internets, the AUP requires companies to be in the
> industry
> and use the network to buy or sell services from each other. This may be
> something like a private club but it is still a public network within that
> industry because you and your competitors are all on the same internet.

That's still proprietary, by my lights. Doesn't really matter whether it
uses IP, baked bean cans and bits of string, or whatever. If you're
enforcing some sort of access policy, then it's private. Period.

Given that the vast majority of companies are moving away from precisely
that sort of scenario to the public Internet, I can't really see people
investing in such a thing - while it's fine for defined circumstances (e.g.
dropping the serial lines to the LSE as part of the Sequence program in the
mid-90s), but I personally can't see it as anything but a policy reversal.
YMMV.

> >That appears to suffice for Reuters, who, if memory serves, have a
> private
> >IP network that encompasses pretty much all large and small players in
> that
> >industry, with a variety of end-points at each customer premises.
>
> Reuters did not use RFC1918 addresses for their network; they simply
> borrowed
> ranges like 1/8, 2/8, 3/8, 4/8 etc. But that's history now. For over three
> years
> Reuters has been migrating their services to larger shared internets which
> use
> registered IP addresses. My company happens to run one of those shared
> internets,
> perhaps the larger one. Our internet also has about 100 providers of
> financial
> services in addition to Reuters on it.

IIRC, Radianz were the Reuters spin-off when they decided to cut costs -
post Bridge and Instinet - and flog HPSN (if I recall the name correctly -
it's been a long time), under the tutelage of Dave Ure (again, if memory
serves).

Reuters themselves moved away from the fixed-connection Monitor network into
IP, and, IIRC, provided DEC PDP-8 based units to allow NAT access to their
(supposedly!) low-latency network by customers. [I say "supposedly" because
a 30-second delay between the old switching centre at Singer Street and
Singapore Editorial was more than a bit inconvenient in a VT320-based
application)]

I would suggest that your target customers would be best served by a
consortium-led directory made up of the main players in the industry, and
that - therefore - this isn't really a valid topic for discussion here. I
personally see it more akin to synchronising POTS telephone directories than
something requiring the formation of an independent international body.
(Obviously, I'm more than willing to listen to other opinions on this).

Please feel free to take this discussion private, should you prefer..

Regards,

Ian Baker
EMEA Support Manager
OpenConnect Systems Ltd.