Re: FYI: Interesting dicussion over at rest-discuss

On 1/10/12 10:36 PM, Markus Lanthaler wrote:
> Great work as always Gregg. I've updated the introduction a bit further so
> that RDF is now mentioned the first time in Advanced Concepts.
>
> Another thing that might help is to change the terminology slightly. We
> currently use subject, predicate, and object which is RDF speak and might
> confuse some people. We could mitigate that by talking about nodes which
> have properties whose values might be other nodes or a literal value. See
> also ISSUE-47 which I've created some time ago.

You can also just use the widely understood and broadly used: Entity, 
Attributes, and Values (EAV). We do this in all of our tools i.e., we 
let you flip between EAV and SPO re. 3-tuples (triples).

RDF simply adds the following to EAV:

1. use of URIs
2. type literals fidelity
3. language tags
4. i18n .

All valuable additions, but not valuable enough for RDF narratives to 
continue to exist modulo any genealogical association with EAV.

>
>
> --
> Markus Lanthaler
> @markuslanthaler
>

Kingsley
>
>
>> -----Original Message-----
>> From: Gregg Kellogg [mailto:gregg@kellogg-assoc.com]
>> Sent: Wednesday, January 11, 2012 4:34 AM
>> To: Kingsley Idehen
>> Cc: public-linked-json@w3.org
>> Subject: Re: FYI: Interesting dicussion over at rest-discuss
>>
>> Thanks for noticing this. I'm preparing the specs to be time-stamped
>> later today, as there's been quite a bit of change since the last
>> version was announced. I went through the syntax document and remove
>> quite a few references to RDF, really trying to leave that to the API
>> document. Of course, there is some reference to formats such as RDFa,
>> Microformats and microdata, but this seems relevant to the desired
>> audience.
>>
>> The only other real mention is in the advanced-concepts section on
>> prefixes, and even there only to reference the CURIE definition.
>>
>> I really don't want people to get hung up on the ability to use JSON-LD
>> for RDF, but focus on JSON-LD as an end in itself. If you might take
>> another pass.
>>
>> Moving the Linked Data section to an appendix might also help the
>> readability, I've also done that.
>>
>> Gregg
>>
>> On Jan 10, 2012, at 10:14 AM, Kingsley Idehen wrote:
>>
>>> On 1/10/12 9:55 AM, Markus Lanthaler wrote:
>>>>> It's the narrative itself, the problem is that from the Semantic
>> Web
>>>>> and
>>>>> W3C side of things an EAV based directed graph that leverages URIs
>> ==
>>>>> RDF. Thus, even when speaking of the aforementioned model (no
>> syntax in
>>>>> mind) they say: RDF. The biggest problem is that when people
>> outside of
>>>>> the Semantic Web and W3C encounter the letters R-D-F they
>> triangulate
>>>>> straight to the RDF/XML and all of its problems.
>>>>>
>>>>> It's a nightmare, to put things mildly :-(
>>>> Maybe moving the theoretical Linked Data definition [1] towards the
>> end of
>>>> the spec and describing the data model based on object oriented
>> programming
>>>> where pointers are IRIs would help to improve the situation
>> slightly. What
>>>> do you think?
>>>>
>>>>
>>>> [1] http://www.json-ld.org/spec/latest/json-ld-syntax/#linked-data
>>>>
>>>>
>>>> --
>>>> Markus Lanthaler
>>>> @markuslanthaler
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>> Markus,
>>>
>>> Linked Data isn't the problem, good effort was put into ensuring the
>> Linked Data == RDF misinformation didn't overshadow JSON-LD. As per
>> usual, the problem is more complex.
>>> Proof:
>>>
>>> 1. open up the json-ld spec page (your URL above)
>>> 2. CTRL-F to perform a search and count of occurrences for pattern:
>> RDF, you get 67 !
>>> Look at the opening paragraph. "RDF" has colonized the blurb-space of
>> this spec, in a nutshell. It's the effects of said colonization that
>> leads people outside the Semantic Web and Linked Data communities to
>> these frustrating conclusions that you've encountered :-(
>>>
>>> --
>>>
>>> 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
>>>
>>>
>>>
>>>
>>>
>>>
>
>
>


-- 

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