W3C home > Mailing lists > Public > public-ldp-wg@w3.org > October 2012

Re: ldp-ISSUE-15 (sharing binary resources and metadata): sharing binary resources and metadata [Linked Data Platform core]

From: Kingsley Idehen <kidehen@openlinksw.com>
Date: Sun, 07 Oct 2012 14:40:49 -0400
Message-ID: <5071CCB1.1030501@openlinksw.com>
To: public-ldp-wg@w3.org
On 10/7/12 1:52 PM, Arnaud Le Hors wrote:
> Sorry Kingsley but I have to side with others on this issue. Resource 
> is undeniably the term that is commonly in use for the web and the W3C 
> is using (see Architecture of the World Wide Web [1] and W3C's index 
> of terms [2]). LDP is not the place to depart from it.
>
> [1] http://www.w3.org/TR/webarch/
> [2] http://www.w3.org/Architecture/Terms.html
>
> I also have to say that I'm puzzled by your claim that "entity" is not 
> overloaded. Looking at W3C's index of terms again I see it already has 
> several possible definitions, none of which are the one you refer to.
>
> But I appreciate your desire to bridge communities and if you were 
> interested in developing a document that explains how the different 
> terminologies relate to one another this would be a valuable 
> contribution. We could reference it from the LDP WG page and maybe 
> make it an annex to the spec. Is this something you would be 
> interested in doing?

Here is a simple example that spells out the problem.

Dret wants to express the following in natural language:

A document can describe something.

Ahok: please note, the statement above has nothing to do with a URI 
resolving to content in a variety of formats.

Dret's quest expressed differently in Turtle notation (I've left out the 
prefix declarations for sake of brevity):

<> <#Describes> <#SomeThing> .

The statement above represents a basic entity relationship held together 
by the predicate: <#Describes> .

The description of <#Describes> is as follows:

<#Describes> <#Type> <#Property>.
<#Describes> <#Domain> <#AnyThing> .
<#Describes> <#Range> <#AnyThing> .
<#Describes> <#Label> "Describes" .
<#Describes> <#Comments> "A predicate that associates an entity of type: 
Document, with a Description Subject." .

# Cross Reference with shared vocabularies or ontologies

<#Type> owl:equivalentProperty rdf:type .
<#Domain> owl:equivalentProperty rdfs:domain .
<#Range> owl:equivalentProperty rdfs:range .
<#Describes> owl:inverseOf wdrs:describedby

According to Dret. and those of you that support his views, for sake of 
being consistent with broken literature, you are implying that:

1. <> -- which denotes a resource .
2. <#Describes> -- denotes a resource .
3. <#SomeThing> -- denotes a resource.

To those that can parse this view point, you are missing the fact nobody 
instinctively assumes a human being is a resource when there is no sense 
of realm partitioning i.e., the Web is an electronic medium distinct 
from the real-world.

All I am saying is that existing literature and basic human instincts 
align much more clearly when literature expresses the fact that:

1. <> -- denotes an entity of type: document .
2. <#Describes> -- denotes a entity of type: predicate.
3. <#SomeThing> -- denotes an entity of type: anything (owl:Thing) 
e.g.., a  person, organization, or any other thing.


Right now (even after all the historic data to the contrary re. 
coherence) you still want to pack all this clarity into the obscurity of 
the overloaded use of "resource" . For what its worth, the content of a 
Web document is the resource (the data). De-referencable URIs (as 
denotation mechanisms) enable triangulation to actual content (the 
resource or data) via indirection. Note, indirection is fundamental 
feature that precedes the Web e.g., its delivered via pointers using 
unary operators "&" and "*" in the 'C' programming language, for instance.

Anyway, I've expressed my point. I am moving on.

Kingsley

>
> Best regards.
> --
> Arnaud  Le Hors - Software Standards Architect - IBM Software Group
>
>
> Kingsley Idehen <kidehen@openlinksw.com> wrote on 10/05/2012 05:05:20 AM:
>
> > From: Kingsley Idehen <kidehen@openlinksw.com>
> > To: Henry Story <henry.story@bblfish.net>,
> > Cc: public-ldp-wg@w3.org
> > Date: 10/05/2012 05:07 AM
> > Subject: Re: ldp-ISSUE-15 (sharing binary resources and metadata):
> > sharing    binary resources and metadata [Linked Data Platform core]
> >
> > On 10/5/12 7:58 AM, Henry Story wrote:
> > > On 5 Oct 2012, at 13:45, Kingsley Idehen <kidehen@openlinksw.com> 
> wrote:
> > >
> > >> On 10/5/12 4:29 AM, Henry Story wrote:
> > >>> On 5 Oct 2012, at 01:22, Ashok Malhotra 
> <ashok.malhotra@oracle.com> wrote:
> > >>>
> > >>>> Kingsley, I agree with Erik.  Resource is the term everyone
> > seems to agree on.
> > >>>> And "entity", too, is overloaded.  For example the "Entity-
> > Relationship model"
> > >>>>
> > >>>> On an earlier point you made, I agree that "denotes" is a good 
> word.
> > >>>> So, a URI denotes a resource, which may have several 
> representations.
> > >>>> All the best, Ashok
> > >>> +1 Let's please stick to vocabulary well understood in the 
> semantic web
> > >>> space.
> > >> Is this the target audience? You know that statement opens up a
> > can of worms, one I am not going to push right now.
> > > The easiest way to steal a jewel from someone would be to pretend
> > to the owner
> > > that it is ugly, and instead give them fake plastic jewellery
> > instead in exchange.
> > >
> > >
> > >>> Debates there have gone on for years, and there is no need to 
> duplicate
> > >>> them here.
> > >> Yes, but not for the reason you espouse. You are making an
> > assumption about the target audience that I think is actually
> > incorrect. I don't believe the semantic web community (whatever that
> > actually means) is the target audience.
> > > I don't think we want silly nomenclature debates here.
> >
> > I don't think I am seeking any kind of *silly nomenclature* debate 
> here,
> > or anywhere else.
> >
> > Again, I am only interested in clarity through terminology that build
> > bridges to other communities. My comments are always loaded and driven
> > by experience across many realms, as I am sure you know by now.
> >
> > My comments and positions aren't hard to find online. Luckily, history
> > is building up so you can easily correlate my positions re. these
> > matters. I am interested in learning from past mistakes and getting 
> them
> > fixed when moving forward. That's it.
> >
> > >
> > >>
> > >>> By all means if someone feels like writing an introductory book
> > >>> for people coming from different traditions into this work, then
> > do it: you'll
> > >>> probably sell a lot of books and make a nice sum.
> > >> Again, you are mistaken about the target audience.
> > >>
> > >> You this this is abobut [LDP] -->[Semantic Web].
> > >>
> > >> I actually believe it's about: [LDP] --> [Rest of the Pragmatic 
> World] .
> > > The semantic web is very pragmatic, I am using it all the time.
> >
> > Not my point.
> >
> > > It has evolved some concepts that are designed to work well with 
> REST, and
> > > it is easy to see that.
> >
> > Not my point of concern.
> >
> > >   There is no need to play into delaying tactics
> > > by trying to please people who will never be pleased whatever you
> > do, however
> > > you explain it.
> >
> > There you make a fundamental mistake. I am not targeting an audience
> > with "R-D-F Reflux Syndrome" I am more interested in a realm of folks
> > that already understand this subject matter, the only thing that
> > confuses them is new terminology disconnected from mainstream 
> literature
> > etc..
> >
> > >
> > >>>   But whatever convention we
> > >>> choose is going to be deemed arbitrary - that is what
> > conventions are: a selection
> > >>> among arbitrary options, in order to facilitate coordination.
> > Using non semantic
> > >>> web or webbish vocabulary is just going to confuse people in the
> > semweb side
> > >>> and people in the other spaces.
> > >> People outside the semantic web community already understand the 
> following:
> > >>
> > >> 1. entities
> > >> 2. entity relationships
> > >> 3. entity relationship semantics
> > >> 4. relations
> > >> 5. relations and state
> > >> 6. identifiers
> > >> 7. name resolution
> > >> 8. structured data representation (via entity relationship graphs)
> > >> 9. data models
> > >> 10. across the wire data serialization formats
> > >> 11. denotation
> > >> 12. connotation
> > >> 13. indirection
> > >> 14. first-order logic
> > >> 12. etc..
> > >
> > > Many of the terms are completely acceptable in the Semantic Web
> > and Linked Data vacabulary.
> >
> > And they are known and understood by many outside the communities you
> > outline.
> >
> > Kingsley
> > >
> > >
> > >>
> > >> Kingsley
> > >>
> > >>
> > >>> Henry
> > >>>
> > >>>> On 10/4/2012 4:06 PM, Kingsley Idehen wrote:
> > >>>>> On 10/4/12 6:49 PM, Wilde, Erik wrote:
> > >>>>>> hello kingsley.
> > >>>>>>
> > >>>>>>> +1
> > >>>>>> thanks!
> > >>>>>>
> > >>>>>>> To something along the following lines:
> > >>>>>>>     The Web can enables *entities* to be *denoted* by any 
> (registered)
> > >>>>>>> URI scheme.
> > >>>>>>>     These entities can be represented by content associated 
> with any
> > >>>>>>> (registered) media type.
> > >>>>>>>     In many cases, applications establish specific (i.e.,
> > typed) relations
> > >>>>>>>     between entities, which can either be under their 
> control, or
> > >>>>>>> controlled by another authority.
> > >>>>>> i'd rather stick with the term "resource", which is well 
> established in
> > >>>>>> many of the core web standards.
> > >>>>> I know you think that's the case, based on material out there.
> > But, its going to change. Resource is an overloaded term.
> > >>>>>
> > >>>>>>   "entity" not so much, so while in the end
> > >>>>>> it's just a different label for the same concept, it is one
> > that i don't
> > >>>>>> want to introduce.
> > >>>>> You aren't really introducing anything, you are realigning
> > with what already exists in literature that precedes the Web [1][2].
> > >>>>>
> > >>>>>>   and i am not quite sure what you think you're getting
> > >>>>>> out of using this different label?
> > >>>>> Clarity is always my fundamental goal, use of existing (pre
> > Web) terminology for the same fundamental concepts so that bridges
> > can be built with other communities en route to a cohesive
> > continuum. Disconnecting existing communities (many of which have
> > long mastered these concepts) via choice of terminology ultimately
> > stifles adoption.
> > >>>>>
> > >>>>>
> > >>>>> Links:
> > >>>>>
> > >>>>> 1. http://en.wikipedia.org/wiki/Entity%E2%80%93relationship_model
> > -- Entity modelling
> > >>>>> 2. 
> http://ontolog.cim3.net/forum/ontolog-forum/2012-07/msg00190.html
> > -- a related discussion on the ontolog forum that actually reached
> > amicable conclusion re. this matter.
> > >>>>>> cheers,
> > >>>>>>
> > >>>>>> dret.
> > >>>>>>
> > >>>>>>
> > >>>>>>
> > >>> Social Web Architect
> > >>> http://bblfish.net/
> > >>>
> > >>
> > >> --
> > >>
> > >> Regards,
> > >>
> > >> Kingsley Idehen
> > >> Founder & CEO
> > >> OpenLink Software
> > >> Company Web: http://www.openlinksw.com <http://www.openlinksw.com/>
> > >> Personal Weblog: http://www.openlinksw.com/blog/~kidehen 
> <http://www.openlinksw.com/blog/%7Ekidehen>
> > >> Twitter/Identi.ca handle: @kidehen
> > >> Google+ Profile: https://plus.google.com/112399767740508618350/about
> > >> LinkedIn Profile: http://www.linkedin.com/in/kidehen
> > >>
> > >>
> > >>
> > >>
> > >>
> > > Social Web Architect
> > > http://bblfish.net/
> > >
> >
> >
> > --
> >
> > Regards,
> >
> > Kingsley Idehen
> > Founder & CEO
> > OpenLink Software
> > Company Web: http://www.openlinksw.com <http://www.openlinksw.com/>
> > Personal Weblog: http://www.openlinksw.com/blog/~kidehen 
> <http://www.openlinksw.com/blog/%7Ekidehen>
> > 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





Received on Sunday, 7 October 2012 18:41:14 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 16:17:32 UTC