Re: Proposal to close ISSUE-19: Adressing more error cases, as is

On 6/5/13 1:13 AM, Erik Wilde wrote:
> hello kingsley.
>
> On 2013-06-04 15:17 , Kingsley Idehen wrote:
>> On the assumption that we both agree there is no such thing as
>> unstructured data:
>> HTTP (Hypertext Transfer Protocol) transfers Data.
>> RDF (Resource Description Framework)  enables you describe and
>> understand Data.
>> RDF based Linked Data enables you describe, understand, and refer to
>> Data by combining the items above.
>
> now that sounds really great, but it really is a bit generic. would 
> you mind spelling out for alexandre and me how a client that has no 
> prior knowledge of LDP will, when it encounters text/turtle LDP 
> resources, be able to figure out which links to follow with which 
> interactions (GET/PUT/POST/PATCH), what to send as request payload, 
> and what that is going to do in terms of LDP protocol semantics? thanks!
>
> cheers,
>
> dret.
>
>

Simple, bearing in mind the assumption that LDP means "Linked Data 
Platform" i.e., the principled approach outlined in TimBL's meme about 
Linked Data applies.

Steps:

1. denote (name or "refer to") entities using HTTP URIs.
2. create a document at a Web Addresses (an HTTP URI/URL) so that entity 
names (from #1) can resolve to the content of these documents.
3. use the RDF model (a W3C standard) to create document content that 
describes the named entities (from #1) i.e., describe these entities 
using RDF statements re., document content.
4. make your entity descriptions richer (and more useful) by adding 
relations that associate them with other entities .

#4 means: keep on adding more RDF statements to your entity description 
as you discover more about the entity you are describing.

The text/turtle matter:

RDF (Resource Description Framework) is a composite comprised of:

1. Model -- First-order logic foundation
2. Syntax -- note, this is the grammar of the model
3. RDF Data Expression Notation -- this is where Turtle comes into play
4. Actual RDF Data Serialization -- Turtle also plays here.

An RDF based system that also supports Linked Data principles MUST 
assume that text/turtle content is either RDF expression (meaning: text 
to be post processed en route to manifesting an RDF graph) or that its 
actually a serialization of an RDF graph).

Every Relation (these are entities denoted by URIs too) in RDF is 
denoted by URI. The Relation entity plays the predicate role in an RDF 
statement (another entity). The predicate is the relationship 
facilitator comprised of a Subject and an Object.

# Turtle Notation (describing entities using a collection of RDF 
statements/relationships) #

<>
a <#Document>;
<#describes> <#StatementX>, <#subject>, <#object> .

<#StatementX>
a <#RDFStatement> ;
<#subject> <>;
<#object> <#Document> ;
<#describedBy> <>.

# Turtle end #

Now, barring any typos, you can cut an paste the above into a document 
and then publish the document to the Web using then  file create, save, 
and share pattern. Share the URL of the document, then based on the 
rules outlined you will be able to follow-your-nose (via HTTP URI 
de-reference) through the graph that manifests.

Everything is related. The only question is how? This is what RDF based 
Linked Data is all about i.e., denote entities using URIs such that URIs 
resolve to their meaning via RDF based content that describes the URI's 
referent.

This is ultimately a game of true or false since each Relation is a 
function. Thus, publishing the Turtle content above is enough to prove 
or disprove my claims. We don't need lengthy wordy arguments for that  :-)

-- 

Regards,

Kingsley Idehen 
Founder & CEO
OpenLink Software
Company Web: http://www.openlinksw.com
Personal Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca handle: @kidehen
Google+ Profile: https://plus.google.com/112399767740508618350/about
LinkedIn Profile: http://www.linkedin.com/in/kidehen