W3C home > Mailing lists > Public > www-ws-arch@w3.org > March 2002

RE: D-AG0009; Semantic Web & Web architecture

From: <michael.mahan@nokia.com>
Date: Sat, 9 Mar 2002 09:46:07 -0500
Message-ID: <5C76D29CD0FA3143896D08BB1743296A072C23@bsebe001.NOE.Nokia.com>
To: <distobj@acm.org>, <Mike.Champion@softwareag-usa.com>
Cc: <www-ws-arch@w3.org>


> 
> > So, how should 9 relate to 11 then? I guess 11 is the Web 
> as it is today, 9
> > relates to the abstract model of the Web promulgated by the W3C?  
> 
> I see 11 as identifying two specific architectural points;
> "distributed" (was this intended to be "decentralized"?), and
> "heterogenous environment" (presumably referring to programming
> languages, operating systems, etc..).  9, by virtue of "align with web
> architecture", covers these, IMO.

In my kickoff of AG0011 I did comment on these 2 offered properties of what it means to be consistent with the current web architecture. I believe there are other design principles of the existing web besides these in order to verify consistency. Goal 11 is all about, IMO, identifying all those significant principles/properties/constraints and then defining some requirements which states that our reference architecture does not contradict property_1, property_2, etc. 

If this is correct, it sounds much different than the main theme of 9, which is to focus on the SW principles: machine understandable semantics, KR, non apriori clients, inferencing, etc. However since 9 is about consistency with what exists (current web) and 11 is about consistency (alignment) with what is envisioned (semantic web), one big overall goal of 'consistency' is understandable. 

That said, I see this as more of a management task. We are trying to get to a set of requirements, and divide and conquer is a good strategy - even if some the chosen boundaries are somewhat arbitrary.

> > Also, is "machine automation" part of our mission?
> 
> The working Web service definition we came up with suggests 
> it is, IMO,
> when it says ".. supports direct interactions with other software
> applications or components ..".
> 
> >  One can agree it's a
> > good thing, and certainly related to whatever we mean when 
> we say that a web
> > service is about the web being processable by machines, not 
> just humans.
> > But we don't want to set the bar too high: web services  
> enable hard-coded
> > programs to communicate via the web (much as they can with 
> DCOM or CORBA
> > over LANs), whereas the semantic web is about enabling more 
> flexible,
> > data-driven, and "intelligent" use of the data on the web, no?
> 
> No.  Hard coded programs can already talk over the Web via HTTP's
> methods.  Only, because every HTTP component exposes the same generic
> interface (GET/PUT/POST, etc..), you don't know the specific type of
> the resource you're dealing with.  The Semantic Web gives you that
> information (to start - it goes well beyond that, of course).
> 
> So ... my suggestion would be one goal.  I could live with two goals
> though - I don't think it's that big a deal.  I'm just 
> raising the topic
> of Web architecture now, because my impression is that many WG members
> have not received much exposure to it.
> 

I still think that the best way to proceed is too acknowledge that the last part of 9: "and the overall existing web architecture" is redundant with 11 and should be removed. I am flexible however and if a quorum feels strongly enough to merge these, then that is OK with me. Maybe one bigger goal of overall consistency with subgoals like AG0006.
 
MikeM
Received on Saturday, 9 March 2002 21:16:44 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 3 July 2007 12:24:56 GMT