Issue-80: POST and discovery - was: a discussion with Alex B.

On 10 Jun 2013, at 17:11, Arnaud Le Hors <lehors@us.ibm.com> wrote:

> Hi Ted, 
> I agree this has to do with discovery. I don't know why you think we decided not to have any discovery mechanism. We have several open issues related to discovery and affordances. 
> I've got a proposal to make that I hope we'll get us moving forward on this.

So my simple answer to the discovery argument is that to use 
the ldp:Container class to do the job. Define an ldp:Container
is the set of things on which a POST will create a new resource 

( and as per ISSUE-79 and  ISSUE-73 it would list all its members 
  with ldp:conains )

As per ISSUE-80 "How does a client know which POST requests creates 
new resource" the answer would be simply

 - the LDPC itself says of itself  { <> a ldp:Container }
 - another resource the client trusted would say of that resource { <ldpc> a ldp:Container }

This does not mean it follows automatically that this is the case. It can happen
that
  1. the ldpc is access controlled
  2. the <ldpc> resource is not an LDPC. It wrongly says of itself that it (or
      some resource wrongly says of another resource that it is an LDPC)
      This would happen if for example someone just dumped an ldpc document 
      on the file system. This is equivalent to someone putting a html form up 
      for a shopping cart that points to the wrong cgi - the human would be mislead
      and confused by the UI.

If you don't do that then you'll have to invent a relation to things can POST to to create 
LDPRs, or you'll have to invent another type of thing to fulfil the same role. So it
would be easier to make things simple here.



> --
> Arnaud  Le Hors - Software Standards Architect - IBM Software Group
> 
> 
> Ted Thibodeau Jr <tthibodeau@openlinksw.com> wrote on 06/10/2013 06:57:50 AM:
> 
> > From: Ted Thibodeau Jr <tthibodeau@openlinksw.com> 
> > To: Arnaud Le Hors/Cupertino/IBM@IBMUS, 
> > Cc: public-ldp-wg@w3.org 
> > Date: 06/10/2013 06:58 AM 
> > Subject: Re: STOP! Re: An IRC discussion with Alexandre Bertails re SSUE-19: 
> > 
> > 
> > On Jun 6, 2013, at 04:11 PM, Arnaud Le Hors wrote:
> > > There is no magic. In one way or another the client needs to 
> > figure out that it is dealing with an LDP server. This can be done 
> > at the HTTP level, at the RDF level, or possibly both. 
> > 
> > In other words...
> > 
> > LDP requires a discovery mechanism.
> > 
> > We've skirted this issue before, in questions about how to know
> > what functionality a server has implemented, and in "implementation
> > specific" functionality, about *how* it has been implemented.
> > 
> > I don't recall any decisions that deliver such discovery mechanism
> > (and think I remember some decisions *not* to do so), but the length
> > of this thread seems a clear justification for such delivery.
> > 
> > Ted
> > 
> > 
> > 
> > 
> > --
> > A: Yes.                      http://www.guckes.net/faq/attribution.html
> > | Q: Are you sure?
> > | | A: Because it reverses the logical flow of conversation.
> > | | | Q: Why is top posting frowned upon?
> > 
> > Ted Thibodeau, Jr.           //               voice +1-781-273-0900 x32
> > Senior Support & Evangelism  //        mailto:tthibodeau@openlinksw.com
> >                              //              http://twitter.com/TallTed
> > OpenLink Software, Inc.      //              http://www.openlinksw.com/
> >          10 Burlington Mall Road, Suite 265, Burlington MA 01803
> >      Weblog   -- http://www.openlinksw.com/blogs/
> >      LinkedIn -- http://www.linkedin.com/company/openlink-software/
> >      Twitter  -- http://twitter.com/OpenLink
> >      Google+  -- http://plus.google.com/100570109519069333827/
> >      Facebook -- http://www.facebook.com/OpenLinkSoftware
> > Universal Data Access, Integration, and Management Technology Providers
> > 
> > 
> > 
> > 
> > 
> > 
> > 

Social Web Architect
http://bblfish.net/