Re: [urn] I-D Action: draft-saintandre-urn-example-00

[Keith Moore <[e-mail]>]

On 01/09/2013 10:48 AM, Julian Reschke wrote:
>>>>
>>>> urn:uuid: is used a lot in practice, and I simply don't see a
>>>> practical problem with it.
>> Whether something works well in practice, and whether something conforms
>> to the intent of a standard, are of course two separate questions.
>>
>> Again, URNs were designed to identify resources.   If people use them
>> for other things, that's not a problem so long as such use doesn't
>> degrade the intended utility of URNs.   It's not like the protocol
>> police are going to chase down the users of UUID URNs and put them in
>> jail if those UUIDs weren't chosen to refer to resources.  And of course
>> you can't tell by looking what is named by a URN  - and that is a
>> feature, not a bug.
>>
>> But just because people find uses for URNs that weren't intended,
>> doesn't mean that the URN standard should be changed to encompass those
>> uses.  This _would_ degrade the utility of URNs.
>>
>> To be clear, there's nothing in principle wrong with a UUID URN. What's
>> wrong is using a UUID URN just because what you need is a unique
>> identifier that doesn't refer to a resource, and you want that unique ID
>> to be some sort of URI.   Yes, it probably does little harm most of the
>> time, but it's still not a good idea to promote the practice.
>> ...
>
> "This specification does not limit the scope of what might be a 
> resource; rather, the term "resource" is used in a general sense for 
> whatever might be identified by a URI." -- 
> <http://greenbytes.de/tech/webdav/rfc3986.html#rfc.section.1.1>
>
> Case closed.
yes, but it didn't anticipate that URNs would be widely used to not name 
resources at all.

again, most such uses do little harm.  but that's not an argument for 
perverting URNs to be just
random numbers instead of resource identifiers.

Keith