Re: [metaDataInURI-31]: Initial draft finding for public review/comme nt.

Hi Norm,

In a message dated 7/10/2003 12:13:48 PM US Mountain Standard Time, 
Norman.Walsh@Sun.COM writes:
> Can you give me an example of "two methods for uniquely identifying a
> thing" where you'd want to choose one of them in the URI?
> 

Sure.
Going back to my car example from a previous email.  I could use
the unique identification of the VIN # which is from the automobile
manufacturer.  Or, if I am a state-government agency, I may prefer
to use the automobile's license plate #.

With all the numerous identification schemes, I am sure there are
better examples.

> | But before you can even broach that you have to decide on whether URIs 
> should 
> | be opaque or not. That is where consensus must first be guaged.
> 
> Handed a random URI about which you know nothing, my position is that
> it is opaque and you've got no business peeking inside it trying to
> guess stuff.
> 

But handed a URI with either a known URI scheme (like <A HREF="mailto:)">mailto:)</A> or
from a known organization (based on a domain name), It is appropriate to
peek into the URL according to either the standards of the scheme or 
standards 
of the sending organization.

IMO, both areas of further URI standardization (scheme, domain) should be 
encouraged.

 - Mike
---------------------------------------------------
Michael C. Daconta
Chief Scientist, APG, McDonald Bradley, Inc.
www.daconta.net