W3C home > Mailing lists > Public > public-ldp-wg@w3.org > January 2013

Re: MKCOL for making collections

From: Henry Story <henry.story@bblfish.net>
Date: Tue, 22 Jan 2013 17:51:04 +0100
Cc: "public-ldp-wg@w3.org" <public-ldp-wg@w3.org>
Message-Id: <7F8285A7-502E-40EB-8CCF-BBCC37892A8A@bblfish.net>
To: "Wilde, Erik" <Erik.Wilde@emc.com>

On 22 Jan 2013, at 17:39, "Wilde, Erik" <Erik.Wilde@emc.com> wrote:

> hello all.
> 
> On 2013-01-22 16:03 , "Henry Story" <henry.story@bblfish.net> wrote:
>> ISSUE-36: Can applications create new containers?
>> We cannot make a collection by POSTing a doument on a collection, since
>> that
>> creates a resource. We therefore would need a different HTTP Method to do
>> this.
> 
> we do not need a different method. POST can do anything you want it to do,
> the media type simply has to say what to POST to interact with the
> resource in a variety of ways. POST is the catch-all for all non-safe
> non-idempotent interactions and has no implied semantics.

Yes, that is why you still have not understood the semantic web, or
linked data, and that is distoring your undestanding of REST. I don't 
want to sound harsh, but this is a key point in the modelling of 
all this.

All contents have semantics. Even Atom can be transformed into RDF 
via GRDDL [1]. So there is no difference at the layer we are speaking 
of between any of the contents. There are different representations 
that are more or less useful for a particular use case
( eg: jpegs is better for picture than RDFXML or Atom )
But otherwise it's graphs all the way down and all the way across.


> 
> personally, i'd advise against MKCOL and other HTTP add-ons other than
> PATCH, since they usually are not very widely supported, and have been
> specified with a rather narrow use case in mind.
> 
> but for now, just as clarification: POST is fine to use for whatever we
> want to use it for; we can decide to use MKCOL, but there's no need to do
> it that way.

The problem is that you require to do a switch on the content type. That
is you are confusing the force of the document-act with the content of the 
action. The force is described in the Verb: GET, PUT, POST, GET, MKCOL, 
PATCH. The content is at the descriptive layer. 

Anyway, if that is too abstract, then consider what happens if 
we then switch to json-ld, or rdf/xml or .... whatever comes up
next. You'll need to change the protocol each time a new format
comes up, because you'll have the content determine the 
meaning of the action, rather than work off the action is what HTTP
calls the HTTP METHOD.


> 
> cheers,
> 
> dret.
> 


[1] http://www.w3.org/TR/grddl/

Social Web Architect
http://bblfish.net/



Received on Tuesday, 22 January 2013 16:51:40 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:11:44 UTC