Re: Web Identity 1.0 -- Draft Spec

On 01/08/2014 12:46 PM, Melvin Carvalho wrote:
>
>
>
> On 8 January 2014 12:27, ☮ elf Pavlik ☮ <perpetual-tripper@wwelves.org
> <mailto:perpetual-tripper@wwelves.org>> wrote:
>
>     On 01/08/2014 02:16 AM, David I. Lehn wrote:
>
>         On Fri, Jan 3, 2014 at 3:00 PM, Manu Sporny
>         <msporny@digitalbazaar.com <mailto:msporny@digitalbazaar.com>>
>         wrote:
>         There are issues to consider when comparing SPARQL, Changeset vocab,
>         JSON Patch, or other systems for triple editing.  SPARQL and the
>         Changeset vocab will be able to edit linked data at a lower triples
>         level.  JSON Patch may require that data be framed into a strict
>         JSON-LD format so that the patch paths make sense.  That could
>         require
>         finishing the framing spec.  If you can read and write the full
>         resource data then just HTTP verbs and a read-modify-write process
>         could work for edits.  I imagine editing data with blank nodes
>         is full
>         of issues too.  And the ease of any approach probably depends on
>         implementation details of the client and server.  It could be a
>         challenge to come up with a good general solution.
>
>
>     especially: *I imagine editing data with blank nodes is full of
>     issues too.*
>
>     looking at:
>     https://web-payments.org/__specs/source/web-identity/#a-__typical-identity
>     <https://web-payments.org/specs/source/web-identity/#a-typical-identity>
>
>     makes me wonder how one updates shippingAddress and citizenship
>     information? especially that very often people can have multiple
>     ones in both of them.
>
>     maybe we can just modify #a-typical-identity example to at least
>     include two elements in shippingAddress and work with that?
>
>
> Best practice according to web axioms is to give anything of
> significance an id.
>
> Citizenship example already has citizenID.
"citizenshipId": "223-93-2422"
currently it doesn't use IRI


> Shipping address can be
> given several ids, like when you have multiple addresses in amazon.
> Ideally you'd want a type = Address in there too, there's probably
> already a vocab for that.
just looked at @context https://w3id.org/identity/v1 which maps
"Address": "schema:Address"


>
> There's also bnode skolemization
>
> http://www.w3.org/wiki/BnodeSkolemization
>
> /genid/ is already registered with IANA by the W3C in a .well-known
> location.
>
> http://www.iana.org/assignments/well-known-uris/well-known-uris.xhtml
yes, Manu also have mentioned it his article: 
http://manu.sporny.org/2013/rdf-identifiers/

Received on Wednesday, 8 January 2014 13:04:48 UTC