W3C home > Mailing lists > Public > public-ldp-wg@w3.org > October 2012

Re: Operations on containers

From: Ashok Malhotra <ashok.malhotra@oracle.com>
Date: Fri, 19 Oct 2012 10:46:46 -0700
Message-ID: <50819206.6030807@oracle.com>
To: James M Snell <jasnell@gmail.com>
CC: Henry Story <henry.story@bblfish.net>, public-ldp-wg@w3.org
James:
At another time, in another context I suggested a new HTTP verb (MGET to get
the metadata for a resource) and was told that there was no way that adding a new HTTP
verb would be accepted.

The LDP position is to use POST/PATCH/DELETE on the container to do what you want to do
with LINK/UNLINK.

I think it's worth listing requirements and asking how they translate to POST/PATCH/DELETE
on the container.  For example:

1. Create a resource and add it to a container
2. Add an existing resource to a container
3. Remove a resource from a container
4. Remove a resource from a container and delete it

Henry has provided some answers but it would be good to have them written down,
perhaps as a table.

All the best, Ashok

On 10/19/2012 9:46 AM, James M Snell wrote:
> FWIW... this is precisely the kind of use case that has led me to explore the re-introduction of the LINK and UNLINK http methods.
>
> http://tools.ietf.org/html/draft-snell-link-method-01
>
> Consider...
>
> 1) Create the item
>
>   PUT /profiles/james HTTP/1.1
>   Host: example.org <http://example.org>
>   ...
>
> 2) Link that item to an existing collection
>
>   LINK /profiles/james HTTP/1.1
>   Host: example.org <http://example.org>
>   Link: <http://example.com/my/friends>; rel="collection"
>
> 3) Link the collection to the item
>
>   LINK /my/friends HTTP/1.1
>   Host: example.com <http://example.com>
>   Link: <http://example.org/profiles/james>; rel="item"
>
> - James
>
>
>
> On Fri, Oct 19, 2012 at 9:38 AM, Henry Story <henry.story@bblfish.net <mailto:henry.story@bblfish.net>> wrote:
>
>
>     On 19 Oct 2012, at 18:32, Ashok Malhotra <ashok.malhotra@oracle.com <mailto:ashok.malhotra@oracle.com>> wrote:
>
>>     If I have another usecase.  I have an, already created, resource R which has a URI and I
>>     want to put it in a container C.  Is that possible
>
>     it should be.
>
>     you put
>
>     <http://mydomain.com/foaf#me> a foaf:Person .
>
>     to
>
>     <http://data.fm/friends>
>
>     then you would have
>
>     <http://mydomain.com/foaf#me> a foaf:Person .
>
>     in
>
>     <http://data.fm/friends>
>
>     if you PUT
>
>     <#me> a foaf:Person
>
>     to
>
>     <http://data.fm/friends>
>
>     Then you'll have that resources  contain
>
>     <http://data.fm/friends#me> a foaf:Person .
>
>
>>     or would the conflation of
>>     the container URI and resource URI prevent that?
>
>     The new resource you PUT to, would now contain a metnion of the resource R.
>
>
>>     Or would the URI of the resource in
>>     the container just point to R which could have a different URI?
>
>     The above should be the case. If it is not then one should open an issue.
>
>     Henry
>
>
>>     All the best, Ashok
>>
>>     On 10/19/2012 9:22 AM, Arnaud Le Hors wrote:
>>>     "Steve Battle" <steve.battle@sysemia.co.uk> <mailto:steve.battle@sysemia.co.uk> wrote on 10/19/2012 08:44:36 AM:
>>>
>>>     > It's my understanding that The Opacity Axiom
>>>     > <http://www.w3.org/DesignIssues/Axioms.html#opaque> applies only to clients
>>>     > attempting to pick apart a URI, rather than to the server.
>>>     >
>>>
>>>     Indeed, but I'm not sure clients could live without knowing the magic involved in this scenario and even more so in the case of creating a resource using PUT.
>>>     --
>>>     Arnaud  Le Hors - Software Standards Architect - IBM Software Group
>
>     Social Web Architect
>     http://bblfish.net/
>
>
Received on Friday, 19 October 2012 17:47:18 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:11:41 UTC