Re: [OK?] Re: SPARQL Protocol: suboptimal examples

On Jan 31, 2006, at 11:36 AM, Dan Connolly wrote:

>> You can register application/prs.kendallclark.notation3 or something
>> like that.
>
> I'm looking into it.
>
>>  If the RDF DAWG wants to use Notation3 in a W3C Technical
>> Report and requires a media type for Notation3 for that, it's the  
>> WG's
>> responsibility to do what's necessary so this can be done in  
>> accordance
>> with RFC 4288 and the TAG's finding on internet media types, both of
>> which seem quite clear that use of unregistered and experimental  
>> media
>> types is discouraged.
>
> I'm sympathetic to that argument.

I'm not. The whole issue is sterile and absurd:

1. The MIME type in question is in an *informative*, not normative  
example. It's not normative and we're not *using* the unregistered  
type. We're showing an example of a protocol interaction where an  
unregistered type is being used. That difference matters, IMO.

2. I suspect there are W3C sanctioned URIs where the content-type  
returned is *precisely* this unregistered MIME type. (If not,  
congrats to the MIT/CSAIL & cwm folks for keeping that strictly  
segregated, and my apologies for the implication otherwise.)

So I'll leave it to others to sort out. I'm not bothered by  
*illustrative* material in a spec, when there are *real* world uses  
of the same unregistered MIME type.

Cheers,
Kendall