Re: rel names vs URLs

On 16 Jul 2014, at 06:25, James M Snell <jasnell@gmail.com> wrote:

> Just an FYI related to this thread:
> 
> I've put together a small script that builds a JSON-LD context from
> the current IANA Link Relations registry.
> 
>  http://linkrels.mybluemix.net/links
> 
> It uses the base URL "http://www.iana.org/assignments/link-relations#"
> for all items. This would obviously only cover registered link
> relations.

I have been creating iana URLs like that too, but I don't think anymore it is such
a good idea.

 The problem is that not being the administrator of iana you can't really control
the meaning of those URLs, so we could not come to an agreement about them here,
without involving iana. If we could get iana to agree that would of course be
completely different, but it probably would be a long process.

The W3C iana urls could be regularly updated by a script that fetches the iana 
information and publishes it locally. It could also contain owl:sameAs relations
and "deprecated" tags for the case of urls that have better known equivalents ( such
as rdf:type )




> 
> - James
> 
> On Thu, Jul 10, 2014 at 4:43 AM, henry.story@bblfish.net
> <henry.story@bblfish.net> wrote:
>> 
>> On 10 Jul 2014, at 00:12, Sandro Hawke <sandro@w3.org> wrote:
>> 
>>> This is only tangentially related to LDP.   I'm sending it here because we've thought/talked about this problem a bit.  I don't think it actually affects the LDP spec, although it might affect LDP users in some situations.
>>> 
>>> My particular use case is I want to use webmention with RDF data, but the webmention folks have no particular interest (I think) in giving me an RDF predicate URL for rel=webmention.     And I'd like to be able to put it in the data, not just in a Link header (for the same reason as the webmention spec allows it as <a rel="webmention"> not just in Link headers.
>>> 
>>> So, I'm wondering about just declaring that http://www.w3.org/ns/rel#X means the HTTP link relation X.   As I understand Web Architecture, that's perfectly within the purview of whoever owns http://www.w3.org/ns/rel.   It wouldn't necessarily be the only URL for the relation; it would just be one easy option.
>>> 
>>> For people who don't want to use URLs for relations, this wouldn't affect them, and they wouldn't need to know about it.  This would NOT mandate that systems which understand rel=X have to also accept rel=http://www.w3.org/ns/rel#X.  It would just allow every RDF system to know one way link relations might be shown in RDF.
>>> 
>>> The only technical issue I can think of -- and this isn't a problem, just a question -- is whether dereferencing http://www.w3.org/ns/rel#type gives you a triple saying it's the same as http://www.w3.org/1999/02/22-rdf-syntax-ns#type.    I'd lean toward Yes.
>> 
>> +1
>> 
>>> 
>>> Does anyone see a TECHNICAL problem with this idea?
>>> 
>>> Who do you think might dislike it for non-technical reasons?    Feel free to respond off-list on this last point.
>>> 
>>>     -- Sandro
>>> 
>>> 
>>> 
>>> 
>>> 
>> 
>> Social Web Architect
>> http://bblfish.net/
>> 
>> 

Social Web Architect
http://bblfish.net/

Received on Wednesday, 16 July 2014 04:38:38 UTC