Re: [RAM] Re: Ramblings about "locator"

This seems wrong to me.
Given any identifier we care to define, someone can later come along 
and decide to "route" to that over some scope.
If that changes the "identifier" into an "address" or a "locator" 
then the terms do not refer to the name but to the combination on the 
name and the way it happens to be used, at the moment.  That seems wrong.

I am driven more and more to a notion along the lines suggested by 
Mr. Lear, that a locator is a naming string (binary or otherwise) 
that changes when you change location, and an identifier is a naming 
string that does not change when one changes location.  (Obviously, 
there are still edge cases.)  Or, to paraphrase Noel's description 
(and to mangle some work from John Day) a locator is where given 
several such things some notions of distance (closeness, relative 
topological placement) can be applied to the entities based on those locators.

Yours,
Joel

At 02:21 PM 6/14/2007, RJ Atkinson wrote:
>>MAC addresses are used as serial numbers in many products. That is
>>an ID if I ever thought there was one. And a MAC address is
>>certainly used to find (that means "where", and "where" means
>>location) an ethernet attached station in a L2 switched network.
>
>Right.  Using my terminology proposal, that makes them "addresses"
>(i.e. objects with mixed semantics, depending on use).

_______________________________________________
RAM mailing list
RAM@iab.org
https://www1.ietf.org/mailman/listinfo/ram