Re: [urn] naming and the necessity of resolution

[Juha Hakala <[e-mail]>]

Hello Lars; all,

Svensson, Lars wrote:

> Since we're looking at the complete family of URN RFCs, I suggest the following:
> 
> RFC 2141bis simply states what URNs are for and what they look like (syntax). It possibly mentions that URNs can be resolvable, but that they don't have to.

OK.

> RFC 3406bis states that when you register a namespace you must decide if it needs resolution services or not. If it needs resolution services, there is a mandatory section where you describe if you will provide services according to rfc2483bis or other services (which).

OK. Please note that RFC2483bis does not try to list all possible 
services (since that is not possible). Instead there should be an IANA 
registry for formal and informal services, and a possibility to create 
experimental resolution services as well.
> 
> If there is agreement on this, I shall supply some text suggestions next week (or possibly the week after).

Great!
> 
> I also volunteer to supply some text to 2483bis (outside of the WG), since we need something to refer to in 3406bis.

Not only do we need something to refer to in 3406bis; 3406bis and 
2483bis must be in synch.

You are most welcome to revise the 2483bis (which will then become a 
joint  venture between you, Alfred and me). I'll send the latest version 
of the document to you separately later today.

> I fear that it is inevitable that some things turn out to be more valid and useful than others. A phlogiston made much sense in the 17th century but it doesn't any more (except to people interested in the history of philosophy, but to them it might still be very important). What I'm trying to say is that it is terribly hard to predict what will make sense in the future and thus we should be very careful not to exclude namespaces just because we don't believe they will be useful in the future. We definitely need some criteria (who will maintain the namespace/resolution service; do they have a plan what will happen, when the organization goes out of business), but also must be able to measure each criterion objectively.

Come to think of it, there is probably nothing much to be afraid of. 
What we need is a flexible system where we have formal, informal & 
experimental namespaces and, independently, formal, informal and 
experimental services. Any combination of a namespace & related services 
is valid, and since it is possible to register the services as needed 
the registrants can indicate precisely what their intentions are. The 
only thing that may be harmful to the URN system as a whole is a 
namespace registration request which fails to provide complete / correct 
information about the namespace.

Best regards,

Juha
-- 

  Juha Hakala
  Senior advisor, standardisation and IT

  The National Library of Finland
  P.O.Box 15 (Unioninkatu 36, room 503), FIN-00014 Helsinki University
  Email [e-mail], tel [phone]