Re: [RAI] Global Service Provider ID - draft-pfautz-service-provider-identifier-urn-01

[Hadriel Kaplan <[e-mail]>]

You kept saying "Million", but I think you mean "Billion". :)

ITADs handle 4.3 _Billion_ numbers.

While there are more than 4.3 Billion E.164 numbers in use, it's in the same ballpark... and it's also the same ballpark for number of devices on the Internet.  

In short: it's a crap-load.  

And this really does make a difference for machines still today - there're still 32-bit CPU's/OS's out there. (and for some reason compilers/libraries still don't seem to handle 64-bit long longs very well)

-hadriel


On Sep 27, 2011, at 10:15 PM, Adam Roach wrote:

> We have dealt with unexpected success in the past, and I'm sure we can deal with it in the future. Jon Postel used to hand out IP address all by himself. Allocation progressed through IANA to ICANN, into a complex delegated system involving regional registries and an insane number of registrars under those registries.
> 
> I really don't think this is the greatest part of our concern. Figuring out whether 4.3 million or 10 million or 18.6 quintillion or 100 quintillion is the correct maximum number of entries is probably relevant.
> 
> Personally, I don't think there's any good reason to second-guess the analysis that TRIP made on this point (they concluded 4.3 million), but I'd love to hear arguments to the contrary.
> 
> /a
> 
> On 9/27/11 20:57, Sep 27, Richard L. Barnes wrote:
>> If you've got someone to delegate it to, it's not unheard of for URN namespaces to get delegated to other organizations.
>> <http://tools.ietf.org/html/rfc5165>
>> 
>> Depending on who it gets assigned to, though, that may change its utility for a general Internet protocol.
>> 
>> --Richard
>> 
>> 
>> On Sep 27, 2011, at 7:39 PM, DRAGE, Keith (Keith) wrote:
>> 
>>> Just to note that issuing 1,000,000 registrations in one year would require IANA to handle something like 10 registrations a minute for every working hour of every working day of the year.
>>> 
>>> That would mean increasing the current staff at least 10 fold, just to handle this load, and require a substantially different financing and operating model.
>>> 
>>> If you think we are ever going to get anything like these numbers, then the IANA registration route is not the way to go. You need a space where you can devolve the allocation to subsidiary parties and so on.
>>> 
>>> Regards
>>> 
>>> Keith
>>> 
>>>> -----Original Message-----
>>>> From: [e-mail] [mailto:[e-mail]] On Behalf Of Dean
>>>> Willis
>>>> Sent: 27 September 2011 18:27
>>>> To: [e-mail]
>>>> Subject: Re: [RAI] Global Service Provider ID - draft-pfautz-service-
>>>> provider-identifier-urn-01
>>>> 
>>>> On 9/27/11 11:19 AM, Richard Shockey wrote:
>>>>> +1
>>>>> 
>>>>> Of all the things to worry about here I doubt this is one of them.
>>>>> 
>>>>> Personally I think folks and enterprises small or large prefer buy
>>>> services
>>>>> from service providers even if the service provider is internal to the
>>>>> enterprise or cloud. Life is too short.
>>>> Having once been in the business of setting up internet connections for
>>>> small technology companies all the way up to big retail/industrial
>>>> firms, I can say that most of them really WANTED to own their IP address
>>>> space so that they could, if they wanted to, dual-home it via two
>>>> connections and advertise the routes via BGP routing adverts. I
>>>> therefore suspect that more of them want to have SPIDs than you might
>>>> expect.
>>>> 
>>>> So while I think that SPID is a Very Important Thing and I'm strongly
>>>> sympathetic to the "just reuse ITAD" argument (and no, the reference to
>>>> "telephone" in ITAD doesn't bother me; all of telephony and all of the
>>>> Internet are semantically merging anyhow), it does seem that we might
>>>> want to think about address space concerns.
>>>> 
>>>> We're at about ten billion entities today -- some "natural persons",
>>>> others "legal persons" like corporations, co-ops, government branches,
>>>> independent business units of a larger entity, ships, factories,
>>>> warehouses, office buildings, and so on.
>>>> 
>>>> With any luck, the population will merely double over the next
>>>> generation. It could be worse. Also with any luck, we're going to need
>>>> comms identifiers for interplanetary and perhaps extrasolar resources.
>>>> Not to mention the "internet of things". How many SPIDs does a swarm of
>>>> von Neumann probes
>>>> (http://en.wikipedia.org/wiki/Self-replicating_machine) require so that
>>>> all of those little ETs can phone home? Just one, or are they going to
>>>> subdomain by stellar cluster?
>>>> 
>>>> While I don't expect every entity to require a SPID, there's an argument
>>>> to be made that the number needed might well exceed 32-bits worth at
>>>> some point in time.
>>>> 
>>>> Consider also the model of "permanent allocation". We don't currently
>>>> reclaim ITAD numbers from failed entities or those who no longer need
>>>> them. Given churn and time, that will eventually eat up a lot of
>>>> numbers. If we only issue a million a year, we burn through
>>>> 4,294,967,296 in a mere 4,294 years. But if we're issuing a hundred
>>>> million a year, we've only got a 42 year window.
>>>> 
>>>> Of course, one could do "reclamation" within a number space. For
>>>> example, the PEN registered to KRAP might, on Hadriel's expiry, be sold
>>>> with his effects. This artificial scarcity potentially creates a
>>>> lucrative market. Consider that New York City taxicab medallions have
>>>> historically been a safer investment than gold; supposedly they have
>>>> never declined in value, and currently sell for somewhere north of
>>>> $600,000 each. I believe that New Mexico state liquor licenses have
>>>> similar properties. If we have a 10% annual reclamation cycle, we can
>>>> issue 4.2 million "recycled" numbers a year in perpetuity.
>>>> 
>>>> But if it isn't our intention to assure a perpetual funding model for
>>>> IANA based on artificial scarcity of numeric resources occurring a few
>>>> hundred years from now, perhaps we'd benefit from a larger address space.
>>>> 
>>>> On the other hand, perhaps we're being wildly optimistic about the
>>>> importance of the SPID and the future demand for them.
>>>> 
>>>> One might also note that 32 bits is a nice fixed-length number in binary
>>>> (32 digits), quaternary (16 digits), or hexadecimal (8 digits). Why are
>>>> we restricting ourselves to 10 decimal digits, which doesn't map to
>>>> anything binary in a very nice way?
>>>> 
>>>> --
>>>> Dean
>>>> _______________________________________________
>>>> RAI mailing list
>>>> [e-mail]
>>>> https://www.ietf.org/mailman/listinfo/rai
>>> _______________________________________________
>>> RAI mailing list
>>> [e-mail]
>>> https://www.ietf.org/mailman/listinfo/rai
>> _______________________________________________
>> RAI mailing list
>> [e-mail]
>> https://www.ietf.org/mailman/listinfo/rai
> 
> _______________________________________________
> RAI mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/rai