[address-policy-wg] 2013-03 New Draft Document and Impact Analysis Published (No Need - Post-Depletion Reality Adjustment and Clean up)

On Sun, Aug 04, 2013 at 03:35:25PM -0400, Tom Vest wrote:
> 
> On Aug 4, 2013, at 10:59 AM, bmanning at vacation.karoshi.com wrote:
> 
> > On Sun, Aug 04, 2013 at 04:35:25PM +0200, Jan Ingvoldstad wrote:
> >> On Sun, Aug 4, 2013 at 3:41 PM, David Farmer <farmer at umn.edu> wrote:
> >> 
> >>> On 8/1/13 12:27 , Tore Anderson wrote:
> >>> 
> >>>> * Nick Hilliard
> >>>> 
> >>>> On 01/08/2013 07:38, Tore Anderson wrote:
> >>>>> 
> >>>>>> +Fair use: Public IPv4 address space must be fairly distributed to the
> >>>>>> End Users operating networks.;
> >>>>>> 
> >>>>> can you define "fair"?
> >>>>> 
> >>> I believe the primary definition of fairness the RIR communities have been
> >>> using is, "only those that have *verified operational need* get Internet
> >>> number resources".
> >> 
> >> This is how Internet number resources have been handled for years;
> >> organizations without verified operational needs have received Internet
> >> number resources, some in huge quantities.
> >> 
> >> One could easily argue that this is one of the root problems with former
> >> Internet number resource handling.
> >> 
> >> Fortunately, IPv6 came to the rescue.
> > 
> > 	Pragmatically, there is zero chance of verification of operational need 
> > 	for anything larger than a /96 in IPv6 space....  So the rules for v6
> > 	allocation actually are fairly close to the original v4 allocation policies.
> > 
> > 	The concept of verified operational need arose in times of scarcity, when 
> > 	there was -no- other option.  Those times have past and its not clear to me
> > 	why there remains this slavish devotion to having an unlicensed regulator
> > 	second guess the viability of a given operational model.
> 
> Hi Bill,  
> 
> Out of curiosity, about how many Class As would you say had been assigned before that scarcity came to be widely recognized, and the
>  concept of "verified operational need" was first articulated? At that time, was promptly doling out all of the remaining Class As on a first first-come-first served basis *not* considered an option? If not, why not? If the prospect of creating an operating environment in which critical number resources were collectively, indefinitely controlled by a relatively small minority of the total number of operators who could such resources to similarly good use (and whose commercial/existential viability would forevermore be contingent on the terms of their access to someone else's number resources) was not considered a viable option back then, what makes it any more palatable now? 
> 
> Granted, both the numerator (resource-haves) and the denominator (have-nots) numbers have grown several orders of magnitude over the ensuing years, but how exactly does that matter if the resulting fraction remains broadly unchanged? 
> 
> Curiously, 
> 
> TV

	/8, pre or post classfull addressing?
	pre-classful, all there was were /8's.   Kind of like all there are now
	are /64s in v6 land.

	when ~10% of the total v4 pool was allocated, scarcity triggered the creation
	of classful addressing, which gave us /16 and /24 space in v4-land.

	at ~40% of the total v4 pool, scarcity created CIDR and started on developing v6-land.

	I'd suggest that with the viablity of IPv6, that the remaining v4 pool is simply 
	vestigal and its getting far more traction than it deserves.   We are clearly
	not in a position of scarcity when we hand out, to individual devices, the functional
	equivalent of the -entire- v4 pool, raised to the 32nd power.

	That is the most profligate waste of address space I have ever seen and these petty
	"verified operational use" arguments seem farcical and almost hypocritical in
	comparison.

/bill