I agree and disagree.
I agree that we should not be attempting to build business models into the
technology. I think that is bad for the technology and bad for business. I
beleive that the email encryption market would be much larger today if folk had
not tried to build in a particular business model as the only one supported.
But I do hink we need to give these issues an airing because the cost of not
doing so is that we end up with unspoken agendas driving decisions, agedas that
would rapidly unravel if given attention.
So for example Keith's latest claim that I am thinking of the benefit of ISP
profits in proposing that virtually every party who is not in the business of
providing access to others is likely to find a /64 ample provision in practice.
I agree that a default allocation of a /128 would create an issue, but I see no
means of extracting functional pricing that is possible with a /64 that is not
possible with a /48.
In fact Keith's proposition that NAT is evil and must be prohibited in IPv6
makes it much easier for ISPs to impose functional pricing as they can charge
according to the number of IPv6 addresses that they see in use.
A long time ago I worked in the control enginering group at ICI. While I am
quite prepared to beleive that in ten years or so I will have 100+ IP
addressable devices in the house I do not for a moment accept that they will be
externally visible. Telemetry data tends to only have value after
interpretation and interpretation is almost always context dependent. Reading
the PH meter in the effulent flow is meaningless unless you know the status of
the three term controller since a high reading when the controller is offline
probably means that someone is washing the PH probe.
So no, I don't think that I will have 100+ Internet accessable devices in the
house even though they have IP. Instead I expect that communication from the
site (i.e. house) will be mediated through a control center in the same manner
as for process control. Further I don't think it very likely that even with the
highest degree of automation imaginable that I would have more control loops in
my house than were necessary to manage a chemical plant in the 1980s,
particularly given that control loops have significant management costs.
Keith's issue is real, his solution is bogus. The only way to ensure that ISPs
do not impose functional pricing is competition. The way to make competition in
the area of broadband provision effective is to comoditize the product by
establishing a set of performance metrics for broadband service delivery.
________________________________
From: Bob Braden [mailto:braden(_at_)ISI(_dot_)EDU]
Sent: Wed 29/08/2007 12:53 PM
To: john-ietf(_at_)jck(_dot_)com; narten(_at_)us(_dot_)ibm(_dot_)com
Cc: ietf(_at_)ietf(_dot_)org
Subject: Re: IPv6 RIR policy [was Re: IPv6 addresses really are scarce afterall]
In this whole discussion, I find it hard to keep separate the
technical issues, about which the IETF should care a lot, from
the business model and issues, about which the IETF should be
agnostic. We may personally care a great deal about the business
issues, but we cannot speak as an organization about them.
Bob Braden
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf