[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]


To: Bill Manning <bmanning@isi.edu>
Cc: michaelm@netsol.com, Hollenbeck Scott <shollenbeck@verisign.com>, ietf-provreg@cafax.se
From: Michael Mealling <michael@bailey.dscga.com>
Date: Mon, 12 Mar 2001 13:47:05 -0500
In-Reply-To: <200103121804.f2CI4Ki17252@zed.isi.edu>; from bmanning@isi.edu on Mon, Mar 12, 2001 at 10:04:20AM -0800
Reply-To: michaelm@netsol.com
Sender: owner-ietf-provreg@cafax.se
User-Agent: Mutt/1.1.2i
Subject: Re: Unique handle generation

On Mon, Mar 12, 2001 at 10:04:20AM -0800, Bill Manning wrote:
> %>% I'm not sure I understand the point you're trying to make in the context of
> %>% admin repository changes over time.
> %>% 
> %>% What happens to your example handle if the admin function is one day moved
> %>% from ARIN to some other administrative entity (call it !ARIN for the sake of
> %>% argument), but the person and prefix remain constant?  Does the handle
> %>% remain WM110-ARIN, or must it change to WM110-!ARIN?
> %> 
> %> 	The legecy composite  WM110-ARIN remains
> %> 	If -I- chose to do business w/ !ARIN, then !ARIN may
> %> 	create the composite WM110-!ARIN.  If !ARIN wishes
> %> 	to create the composite anyway, they -MUST- use my
> %> 	handle, WM110, not create something else.. like BM14
> %> 	or WM73.
> %
> % So your personal handle is part of a globally unique space that all
> % registries have to coordinate to ensure uniqueness, right?
> % 
> % -MM
> 
> 	Unfortunately, yes.
> 	Just like the registry handles have be be globally unique.
> 	(It would be confusing to have two registries with the same
> 	handle... :)
> 	One must have globally unique space within each object type.

So, you have three methods that I'm aware of to do this:

1) all of the registries have to cooperate to make sure they are
   creating unique handles 

2) some other entity has to give out unique object handles.

3) come up with a scheme that can be self assigned but still
   gaurantees uniqueness (some hash function, uuid being just one)

I think the point that others are attempting to get to by
suggesting that an object's name is its registry handle 
plus its handle is to avoid these methods.  It would be nice of 
folks said why they find these methods objectionable..

-MM

-- 
--------------------------------------------------------------------------------
Michael Mealling	|      Vote Libertarian!       | www.rwhois.net/michael
Sr. Research Engineer   |   www.ga.lp.org/gwinnett     | ICQ#:         14198821
Network Solutions	|          www.lp.org          |  michaelm@netsol.com

Home | Date list | Subject list