W3C home > Mailing lists > Public > public-sweo-ig@w3.org > November 2006

Re: Semantic Web Layer Cake Update Suggestion

From: Martin Dzbor <M.Dzbor@open.ac.uk>
Date: Thu, 30 Nov 2006 18:23:37 -0000
Message-ID: <015a01c714ac$a78b8ea0$2f186c89@open.ac.uk>
To: "W3C SWEO IG" <public-sweo-ig@w3.org>

Hi all,

with a slightly different take on, may I suggest treating the "application,
UI, services, etc. layer" as something independent of the rest of the cake?
IMHO, any layer can be applied or interfaced per-se, or *an* application may
possibly exist at any step in the layer cake... Maybe one way how this could
be expressed is to add an orthogonal dimension rather than a layer, and then
we can say e.g. things like:

"application purely at the layer of URIs/URLs = e.g. web site.... (most of
'web 1.0'?)"

"application at XML level subsumes URI level = e.g. EDI or suchlike"

"app at RDF level = tagging, RSS in RDF........ (some 'web 2.0' stuff?)"
etc. etc.


What this might enable us to do is to attempt some fuzzy boundaries on the
application axis where SemWeb benefits may become more obvious (e.g. at the
level of min RDF-S + SPARQL); at the same time this is a more inclusive view
that does not exclude those people/companies for whom the XML box might be
sufficient to consider their app "semantic"...  So in principle, I would see
the data provision layer as without any bells and whistles, anything from
the cake above it brings the app closer and closer to the full-scale
potential of SemWeb. e.g SQL access is outside of the cake, so no chance
there, but XML endpoint is a tiny step closer to semantics, RDF endpoint
even closer, etc.

IMHO, such a more inclusive view may be easier to reach out with, than
anything implying many layers that are too ill-defined for many
stakeholders... anyway, just a thought - hopefully not too philosophical...
:-)

Regards,

M.
--
Martin Dzbor  --  KMi (The Open University, UK)
http://kmi.open.ac.uk
http://www.neon-project.org





----- Original Message ----- 
From: "Wilson, MD (Michael)"
Sent: 30 November 2006 17:46
Subject: RE: Semantic Web Layer Cake Update Suggestion



Jeff

The main problem with that diagram is the User Interface & Applications
layer at the top which suggests that applications can only be built on
top of the Trust layer, and not on top of each layer.


Michael Wilson
Manager, UK & Ireland Office of W3C
e-Science Centre
CCLRC Rutherford Appleton Laboratory
Chilton, Didcot, Oxon, OX11 0QX, UK
http://www.w3c.rl.ac.uk/ukofficecontact.html
Fax: +44 1235 445831

-----------------------------------------------------------------

The information contained in this message is confidential and intended
only for the individual named above. If you are not the intended
recipient, or responsible for delivering it to the intended recipient
you are hereby notified that any dissemination, distribution, copying,
or disclosure of this information is prohibited and may be unlawful.


-----Original Message-----
From: public-sweo-ig-request@w3.org
[mailto:public-sweo-ig-request@w3.org] On Behalf Of Jeff Pollock
Sent: 30 November 2006 17:30
To: 'Kingsley Idehen'; 'W3C SWEO IG'
Subject: RE: Semantic Web Layer Cake Update Suggestion


Kingsley-

The diagram Ivan sent here:
http://www.w3.org/2006/07/layerCake-4.png

Implies to me that RDF needn't use XML.  Although I don't personally
feel that the XML serialization of RDF is such a big market issue, real
or perceived, the picture there seems to address the concern quite
nicely.

Warm Regards,

-Jeff-


-----Original Message-----
From: public-sweo-ig-request@w3.org
[mailto:public-sweo-ig-request@w3.org]
On Behalf Of Kingsley Idehen
Sent: Thursday, November 30, 2006 9:18 AM
To: 'W3C SWEO IG'
Subject: Semantic Web Layer Cake Update Suggestion


All,

My initial attempt at a revised Cake is at:
http://myopenlink.net:8890/DAV/home/kidehen/gallery/my_photos/Semantic_W
eb_S
tack_Rev1.png

The focus right now is fixing the XML placement issues such that RDF the
Data Model and its optional XML serialization format are distinct.

I have also added a Data Provider / Data Source layer since the URI/IRIs
are "Resource Identifiers" for  Data which has to come from somewhere
:-)

-- 


Regards,

Kingsley Idehen       Weblog: http://www.openlinksw.com/blog/~kidehen
President & CEO
OpenLink Software     Web: http://www.openlinksw.com
Received on Thursday, 30 November 2006 18:23:54 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:17:34 GMT