Re: What is Oracle's objection to the use of Turtle as R2RML syntax?

On 12/14/11 6:04 PM, Enrico Franconi wrote:
> On 14 Dec 2011, at 18:12, Kingsley Idehen wrote:
>>> I do believe that the cause of the non widespread adoption of OWL is *not* because of the multiplicity of syntaxes.
>> Syntax multiplicity isn't the defining issue re. OWL. A specific syntax is, and that's RDF/XML. And if you are interested in history you'll notice RDF/XML has had the same effect on other aspects of the Semantic Web Project. Basically, every time its moved away from the "front door" appreciation, comprehension, and momentum follows.
> mmhhhh. My first paper on description logics dates back in 1989. In 1992 I was part of the group which chose the standard syntax for DLs at the time (it was called KRSS); the same year I had the first serious system paper on DLs. I started working with Dieter and Ian in the late 90ies on defining the syntax of what at the time was still called OIL (later DAML+OIL, and then OWL); I stopped almost immediately horrified by this useless game to encode a logic in RDF -- a game that it is still going on...
> So, I know history and I don't like RDF/XML either.

Fine, but I haven't questioned your knowledge of history.
> However, from my long standing experience, I am personally convinced that the issue of the syntax is playing the role of the scapegoat in hiding the real reasons for the alleged 'failure' of OWL or RDF itself.

Again, OWL hasn't failed. The issue is comprehension, appreciation, and 
exploitation inertia.

> There is, in my opinion, the incompetence of (large part of) the theory people in looking at the real needs for applications (without trusting blindly to practitioners), the incompetence of (large part of) the practitioners in really understanding the added values of languages such as OWL or even RDF, the incompetence of  (large part of) the system people who are mainly interested in pushing their technology regardless of the theory or the real needs.

Yes, there are lots I agree with there. But, with regards to this R2RML 
matter, Turtle as normative syntax is the shortest path to success on 
two fronts:

1. closing work group item
2. evangelism, education, and bootstrap.

>   Even with the best syntax in the world, you do not overcome these issues.

Yes, for the most part. That said, RDF/XML has been responsible for a 
lot of Semantic Web Project related inertia. This isn't really a matter 
folks debate anymore. I refer to Turtle,  SPARQL, DBpedia, and the 
Linked Open Data projects (repeatedly) because each exemplifies Semantic 
Web Project output that gained traction modulo RDF/XML visibility and 
emphasis.

> Look at SQL and RDBs: horrible syntax (nobody can deny that), but it has conquered the world.

No, its advanced the world to a point where other aspects of the 
relational model are now appreciated due to shortcomings in conventional 
SQL implementations. By this I mean:

1. late and loose schema binding or what we call schema last
2. object-relation-object 3-tuples
3. intensional object identity vs. extensional object identity
4. inference and reasoning
5. conceptual (as opposed to application logic) oriented data access .

>   Why? There is real beef beyond this technology, and theoreticians funded companies in the 80ies which did solve real problems.

Yes, and today the same is happening in the realm of Linked Data and 
relational property graph oriented DBMS technology. Especially now that 
we can negate impedance mismatches of yore via effective bridging and 
advances is physical layer DBMS tech e.g., column stores.

>
>>> The above statements remain just your unsubstantiated opinions of yours.
>> Same applies to you. We are both entitled to our subjective opinions :-)
> My point exactly :-)
> But please present your opinions as such, not as definitive statements.

I only know how to make definitive statements knowing that they are 
inherently subjective :-)

Kingsley
>
> On 14 Dec 2011, at 18:51, Richard Cyganiak wrote:
>
>>> I do believe that the cause of the non widespread adoption of OWL is *not* because of the multiplicity of syntaxes.
>> The snytax-level issue with OWL is not the multiplicity of syntaxes. It's that the normative syntax, RDF/XML, obscures the structure of OWL and is universally detested.
> See above :-)
>
> cheers
> --e.
>


-- 

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