[LINK] Enum musings

[Chirgwin, Richard]

David,

We're not at odds on (b); I said administration of the >hierarchy< not of
the delegated databases. Accepting the robustness of the DNS "system"*, the
question is then "who controls the control?"

As observed by Craig McTaggart of the University of Toronto's Faculty of
Law:
http://arxiv.org/pdf/cs/0109091
>The question of who should hold what authority (if any) 
>over  which  elements  of  the  Internet's  logical  infrastructure  is
not
>yet completely  settled.   Management responsibility and limited 
>administrative authority over some Internet naming, numbering, and 
>addressing resources  have  been  delegated  (perhaps  improperly  as  a  
>matter  of  U.S.  administrative  and/or constitutional law . see Section
5(c) 
>below) to the Internet Corporation for Assigned Names and Numbers 
>(ICANN).  However, despite assurances that it would divest itself of such 
>power, the U.S. government continues to hold residual policy authority 
>over these resources. 

Internet people generally tell the world that telephony is hidebound. Maybe
so; but its policy authority is hidebound partly because the ITU was created
by treaty rather than by "ergo sum". 

*Re DNS system-level robustness: a single end user can be isolated from the
Internet by the temporary invisibility of a DNS; so while it's robust at a
macro level, it can misbehave at the micro level. OTOH, all phone calls on
the Telstra network use the same SS7 database; it may be theoretically less
robust at a macro level (if it fails, everybody fails, but that's rare), but
at the micro level it looks more robust (if dialtone exists, I can nearly
always dial the number).

> On point (a):
> The information is about phone numbers, not individuals.  As
> you pointed out, an individual can have multiple phone numbers
> and a phone number can have multiple individuals.

The databases will contain names; from the ACA's discussion paper:
"...all the communications services associated with each telephone number
would be stored in a generally accessible DNS database..." (Page 8, Privacy
and authentication); and "The information may then be used for spamming or
to assume someone else's identity". 

This clearly rests on the assumption that individuals are identified, if
merely to identify the individual as the licensee of the database entry.

Moreover, (IMO) some conflicts cannot be resolved without identity data.
>From the ACA again:
"For example, if geographic telephone numbers are added to an Enum database
in Australia, several issues arise including:
* which member of the household can claim use of the geographic number for
Enum purposes?"
Some database is needed to resolve the individual to the number. If this is
not in the DNS, it must reside somewhere else - leading back to the
proliferation of databases; and risking the name/number map migrating to an
unregulated sphere (for eg, spam is illegal but how do you stop e-mail
address harvesting?).

(PS, thanks to all Linkers for their contributions to this discussion and,
therefore, to my education!)

Richard C

> -----Original Message-----
> From: David Keegel [mailto:djk@cyber.com.au]
> Sent: Wednesday, 9 October 2002 14:50
> To: Chirgwin, Richard
> Cc: 'link@www.anu.edu.au'
> Subject: Re: [LINK] Enum musings
> 
> 
> ] My understanding - subject to correction! - is that the DNS 
> hierarchy is
> ] ultimately tributary to the highest-level directory. So 
> Enum says two
> ] things:
> ] a) create a database in which an individual is identifiable by a DNS
> ] registration; and
> ] b) centralise administration of the hierarchy to ICANN. 
> 
> On point (b): 
> I totally disageee.
> 
> If you wanted to centralise administration, you wouldn't use DNS
> as a vehicle.  DNS is deliberately designed for de-centralised
> administration, and it makes de-centralisation (delegation) easy.
> 
> I thought the idea was that ICANN would delegate e164.arpa to
> someone like ITU or RIPE and the ITU would delegate country-code
> sized chunks like 1.6.e164.arpa to authorities in that country*
> like the ACA.  (* - Let's pretend that North America is a country.)
> 
> This means ICANN would have almost no practical control over
> day-to-day things within ENUM - ultimately all it could do is
> take the whole e164.arpa and give it to someone else rather
> than the ITU or whoever.
> 
> On point (a):
> The information is about phone numbers, not individuals.  As
> you pointed out, an individual can have multiple phone numbers
> and a phone number can have multiple individuals.
> 
> Some of you may not be familiar with Telstra's 0500 (One Number)
> service.  An 0500 number is not connected to a physical phone
> line, but can be diverted to wherever you like (including
> having pre-programmed time-based diversions).
> 
> This can be a useful option for businesses and such which want
> to be contactable or on-call 24x7 (when the office is closed).
> It's not mandatory.  It does mean allocating a number without a
> phone line.  It has been around for years.  I have used it.
> 
> ______________________________________________________________
> ____________
>  David Keegel <djk@cyber.com.au>  URL: 
> http://www.cyber.com.au/users/djk/
> Cybersource P/L: Linux/Unix Systems Administration 
> Consulting/Contracting
> 
----------
For Link list information see http://sunsite.anu.edu.au/link/