[foaf-protocols] WebID spec: supported RDF formats (bis)

Kingsley Idehen kidehen at openlinksw.com
Mon Jul 19 18:35:01 CEST 2010


Seth Russell wrote:
> 2010/7/18 Kingsley Idehen <kidehen at openlinksw.com 
> <mailto:kidehen at openlinksw.com>>
>
>     Jiří Procházka wrote:
>     > On 07/18/2010 07:17 PM, Seth Russell wrote:
>     >
>     >> On Sun, Jul 18, 2010 at 10:07 AM, Nathan <nathan at webr3.org
>     <mailto:nathan at webr3.org>
>
>     > >> Well maybe that attitude is why things don't  usually work
>     all that very
>     >> well in the LinkedData community.   If you want to do it that
>     way, then why
>     >> even write a standards document .,.. just keep playing in your
>     sandbox.
>     >> If we specify conforming agents, then we should be able to rely
>     on them and
>     >> just perhaps things will be more likely to actually work.
>     >>
>     >
>     > Now it's my turn:
>     > +100
>     >
>     -100
>
>     Here's why:
>
>     But how can anyone seriously state that Things Don't Work Very Well in
>     the Linked Data Community. That is exhibit #1 for RDF == Linked Data
>     conflation driven confusion.
>
>
>  
>
>
>     Things are working wonderfully well in the Linked Data community. It's
>     actually THE reason why we have a WebID Protocol taking shape
>     right now.
>
>     And in a few months time we'll be saying: Ah! There's a Read Write Web
>     and its all happened because we have a WebID protocol for Unambiguous
>     and Verifiable Names covering Agents of all types.
>
>     When RDF comes into the mix, confusion occurs (see recent literal
>     subjects imbroglio).
>
>     Seth:
>     You were referring to RDF not Linked Data in your comments.
>     Imagine the
>     Semantic Web Project without Linked Data. 
>
>
> Sorry i don't really understand the distinction.  Could you enlighten 
> me or point me to where the distinction is made in detail ?

Well best you read the Data 3.0 manifesto.

The Semantics of Linked Data are very basic as per the 4 original rules 
in TimBL's Linked Data meme.

Post bootstrap success RDF and SPARQL evolved from implementation 
details (or suggestions) to being mandatory (an controversial 
development that I don't agree with at all, even though we implement 
Linked Data using RDF and SPARQL etc.. at OpenLink).

>
> Let me clarify what i meant by "... things don't  usually work all 
> that very  well in the LinkedData community".   Take for example 
> foaf.  I've watched foaf since it started last century.  Back then i 
> did some studies with  a home grown RDF browser to see if i could find 
> some usable information and whether we could grock a community growing 
> around it.  

FOAF is more of an RDF showcase than a Linked Data meme showcase. 
Remember, it was full of Blank Nodes. In the Linked Data realm, blank 
nodes are discouraged (quite strongly).

Linked Data has made FOAF virtues clearer, and WebID may actually be the 
culmination point of FOAF potential fused with Linked Data etc..

> Well i didn't come up with anything then.   I visited the project 
> again this year and things really hadn't changed. 

See comment above.

FOAF is not your Linked Data baseline example.

DBpedia is your Linked Data baseline example. And from there the rest of 
the LOD cloud that's developed around it plus the best practices guides 
for publishing Linked Data.


> Here is another example:  The US government made a big thing about how 
> open data would be published about stimulus package projects.  Well i 
> went looking for what was happening in my area.    Sorry, nothing 
> there that i could find in the linked data arena.  

Governments efforts are no on par with DBpedia or the rest of the core 
LOD Cloud when it comes to showcase examples of Linked Data. These 
projects are still mastering many of the subtle nuances.

> I actually want to know who are the contractors working in Renton 
> Washington on stimulus projects.   Can i find that in linked data 
> now.  Sorry i can't, but maybe i just havent tried hard enough.   Now 
> i'm going to make a embarrassing confession:  I have, through no lack 
> of trying, never found any information on the Semantic Web  that i 
> have ever used.

Neither have I .

I've found stuff serendipitously though, via the burgeoning Web of 
Linked Data :-)

I don't see the Web of Linked Data as being equivalent to a Web of 
Semantically Linked Data, I say this because the Semantics of Linked 
Data are low fidelity at best.


>
> Now on the other side of the matter i have successfully used triple 
> (and quad) stores in almost every project that i have implement  since 
> the late 70's.   It is the way to store and access data.  I am a true 
> believer. 

You are chanelling E-A-V model here :-)
>
> So what i am looking for are successful projects that actually get 
> useful information into the hands of consumers. 

DBpedia + GoodRelations based Linked Open eCommerce style Data Spaces 
are taking shape. You can FIND interesting things at: 
<http://lod.openlinksw.com>, even more so when we upgrade the UI (since 
nobody bothered to use the REST interfaces for the DBMS hosted faceted 
navigation/pivot engine).

> I really don't think that has started to happen yet  on the semantic 
> web.   What is happening instead is that web services are delivering 
> useful information to consumers every day; but we don't really know 
> what data lies behind those apps,  and that data is not by and large 
> being shared service to service or consumer to consumer.   Yes, 
> potentially it can be shared, there are academic examples galore   ... 
> but in the real world of consumer data i personally haven't  seen 
> anything.  
>
> I  want to experience this thing really working.  For that to happen, 
> imho,  we need to start making some hard decisions, start foreclosing 
> on some of the infinite possibilities of doing stuff, and stop leaving 
> everything in the world up for grabs. 

Linked Data and the Linked Open Data Cloud deliver on that goal.

>  
>
>     I can assure you Google would
>     be sitting cosy rather than scrambling to acquire Metaweb. In
>     addition,
>     if you read between the lines of the Google move (read the press
>     release
>     closely) they blatantly admit: Disambiguation is Hard, so we went and
>     bought a solution to get us some breathing space, we know
>     Disambiguation
>     and navigation across Entity Relationship graphs is something we just
>     don't have an answer for right now etc...
>
>     Google has been involved with the Semantic Web Project in varying
>     capacities since inception (they emply R. Guha who is one of the RDF
>     creators), but only through the success of Linked Data has opportunity
>     cost of doing nothing has become palpable to their business side
>     decision makers.
>
>
> Don't get me started with Google and their "Semantic Web Project" .......

LOL!
>
>     Kingsley
>
>
> Thanks for the dialogue.  OpenLink  software, for me, holds the best 
> prospect of making this work.   But i don't want to just imagine 
> something  working,  instead i want to get actual useful information.    

OpenLink is about "Making Technology Work For You!". We are engaged, and 
if you look closer at the LOD cloud you we see we are a little more 
engaged that is typically obvious.

Just keep watching:

1. http://lod.openlinksw.com
2. http://uriburner.com
3. http://uriburner.com/fct
4. http://del.icio.us/kidehen  -- lots of practical real world usage 
examples


Kingsley
>
> Seth Russell
> Alpha testing: tagtalking.net <http://tagtalking.net>
> Facebook ing: facebook.com/russell.seth <http://facebook.com/russell.seth>
> Twitter ing: twitter.com/SethRussell <http://twitter.com/SethRussell>
> Blogging: fastblogit.com/seth/ <http://fastblogit.com/seth/>
> Catalog selling: www.speaktomecatalog.com 
> <http://www.speaktomecatalog.com>
> Google profile: google.com/profiles/russell.seth 
> <http://google.com/profiles/russell.seth>
>
>


-- 

Regards,

Kingsley Idehen	      
President & CEO 
OpenLink Software     
Web: http://www.openlinksw.com
Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca: kidehen 







More information about the foaf-protocols mailing list