Re: A modest attempt to re-open ISSUE-20

On Thu, 2013-03-14 at 09:57 -0700, Erik Wilde wrote:
> hello james.
> 
> On 2013-03-14 9:52 , James Leigh wrote:
> > On Thu, 2013-03-14 at 09:43 -0700, Erik Wilde wrote:
> >> it's not the client's thing to decide on URIs for new resources, at
> >> least in the usual case. servers manage the URI namespace and assign
> >> URIs for new resources. clients may suggest URIs (well, mostly certain
> >> path components of them) using HTTP Slug, but they cannot depend on the
> >> server actually using these components.
> > Sure, but there is nothing wrong with the client suggesting the URI to
> > use in the posted RDF model itself. The server is under no obligation to
> > accept the RDF as-is.
> 
> you're right that the server could rewrite the request RDF (maybe based 
> on the Slug URI also provided) and thus transform the suggested RDF into 
> the actual RDF being persisted. i think this works on a technical level, 
> but i think it would be a rather odd way of handling things. but that's 
> certainly just my personal view.
> 


I use to think the same thing; I started with the approach currently in
the spec draft and found out the hard way, that it brakes in most
tooling. I hope to save the ldp wg from going through the same troubles
I did.

James

Received on Thursday, 14 March 2013 17:13:13 UTC