Re: skirting hr14

On 21 July 2014 19:23, Sandro Hawke <sandro@w3.org> wrote:

>  On 07/21/2014 12:59 PM, Melvin Carvalho wrote:
> > However, the one area which I dont think is appropriate to 'bending' is
> HR14, which is essentially bending AWWW.  I cant fathom how you are going
> to get interoperability out of this compromise, in the same way that 5 star
> linked data is interoperable.
>
> I think it's possible to sidestep httpRange-14 without violating anything
> in AWWW or RDF.
>
> For a somewhat detailed look at the general approach, see [Simplied
> RDF][1].  I don't think think we need to go quite that far, though.
>
> And we certainly don't need to conflate people and their profile pages. We
> just need to say that we'll refer to people by referring to their profile
> pages.
>
> For example, these are each URLs for sites/profiles of mine:
>
> http://hawke.org/sandro
> http://www.w3.org/People/Sandro/
> https://www.facebook.com/sandro.hawke
> https://plus.google.com/+SandroHawke/posts
> https://github.com/sandhawke
>
>
> The way I think WebID should work, any of them could function as a WebID
> for me for me now, today, with no need for any coding or bytes to
> change.    Hopefully many of them will be RDF Sources with good
> information, and many of them will provide for webapp logins, like we see
> with OpenID or WebID-TLS.
>
> To get my name, if it is providing it, being an RDF Source....
>
> If you get triples from http://www.w3.org/People/Sandro/ you should NOT
> see
>
> <> foaf:name "Sandro Hawke"     ## ERROR!  DONT DO THIS!
>
>  which would be a range violation, indeed.  Instead you would get
>
> <> eg:profileOwnersName "Sandro Hawke".
>
> or
>
> <> a foaf:PersonalProfileDocument;
>       foaf:PrimaryTopic [ foaf:name "Sandro Hawke" ].
>
>
> The main point is that in all the interfaces, all the drag and drop, all
> the permissions, etc, we use Personal Profile Document URLs, instead of the
> direct "person" IRIs, to identify people.
>

I agree.

I'm not anti HR14, im pro interoperability.

Sometimes on the web you can do things "wrong" and things wont break e.g.
<br/> vs <br> in HTML / XHTML -- not a problem.

I think the best solution is to add #this to URIs that you think have made
a mistake, as a kind of auto correct.

Will work most times, but there's this interesting grey area where you do
it when you should not have, or you dont do it when you should have ...


>
>          -- Sandro
>
>  [1]: http://decentralyze.com/2010/11/10/simplified-rdf/
>