[all, wrld] World Task Force issues

I was ACTIONed to explain why I think putting the "world" task force 
on hold for now is a good idea - here is my rationale.

  To start, let me make clear what I think the role of this Working 
Group is - but let me make clear that this is my opinion and in no 
way represents a position as either an AC or a member of the W3C Sem 
Web Coordination Group.
  Working Notes from this Working Group, it seems to me, should 
represent consensus on  issues that have more to do with design than 
evangelism.  A wonderful example of this is the Turtle stuff, in 
which we are able to endorse a widely used "presentation syntax" for 
RDF/OWL and give it a status above what could be done by an 
individual publishing to an interest group or even an organization 
presenting this as a W3C Note.  Our endorsement says, in essence, 
that a number of W3C organizations have had their eyes on this 
design, believe it is important to the use of the Semantic Web, and 
promote its continued use.
  An example of what we shouldn't do, it seems to me, is anything 
whose purpose is primarily "evangelism" of the Semantic Web.  There 
is a Semantic Web Education and Outreach component to the W3C 
Semantic Web activity, and while this WG should be interacting with 
that, I don't see doing this sort of thing as the right role for a WG.

So, with that in mind, let's look at the two things I hoped the World 
Task Force would do:

  1. explain how RDF interacts with other web architecture components 
and languages    -- what I had in mind was something on the technical 
level, not a "RDF is wonderful" document, but a "here's how to make 
it interoperate with other stuff" note.  What I have found by 
watching the GRDDL work is that this is very hard to do right, and I 
think we should do it fairly specifically in the context of one thing 
at a time (for example, I'd love to see a standard XSD to RDFS/OWL 
conversion - my group encounters this  all the time, and it would be 
nice to have some "standard" so we could interoperate w/other people 
doing the same).  I think pursuing these individual cases and 
pursuing them through notes (and even Rec documents if they are 
really important, as GRDDL might be) is a valuable thing for the WG 
to do - but I'm convinced a general statement would be much work for 
little gain.  I'd strongly urge everyone on this group to write 
papers evangelizing the great things you are doing with Semantic Web 
languages, but I think these will have more impact in XML.com, XML 
Journal, large workshops and conferences, and other venues where this 
sort of outreach and evangelism is more appropriate

  ii. explain the whole continuum of RDF/RDFS/OWL{Lite,DL, Full} -- I 
still think this is important, and eventually needs doing, but 
frankly I simply don't believe we know enough yet and if we produced 
a bad note on this I think it would do much more damage than to 
produce nothing.  On the DAWG, for example, I find that another 
member and I seem to diagree violently every time OWL is discussed - 
the interesting part is that he is from a startup basing its survival 
on OWL and, like me, is a big OWL promoter -- so the main 
disagreement is between two OWL promoters.   Similarly, I don't see 
how we could come to agreement on a discussion of the costs and 
benefits of OWL Lite v. DL v. Full in anything other than theoretical 
terms until a lot more applications and implementations are out 
there.   So, basically, if I thought there was agreement on the WG I 
might be willing to see if we could move this forward, but I don't 
see major points of agreement emerging (look how hard it is to do 
some of the "simple" things in OEP in this consensus process) and I 
think we would be introducing a very contentious issue when, frankly, 
I simply don't think we know the answers.

So, based on the above, my opinion is that World would detract from 
other, IMO more important work, in these early stages of the WG.  I 
believe that in another year or two, based on what I am seeing in 
uptake of RDFS/OWL, I think we will know more and can at least argue 
from practical, rather than theoretical, perspectives, and I think 
that we can reconsider the RDFS/OWL(etc) issues at that time

  Not sure if that helps, but it reflects why I made the recommendation I did.
  JH


-- 
Professor James Hendler			  http://www.cs.umd.edu/users/hendler
Director, Semantic Web and Agent Technologies	  301-405-2696
Maryland Information and Network Dynamics Lab.	  301-405-6707 (Fax)
Univ of Maryland, College Park, MD 20742	  240-277-3388 (Cell)

Received on Wednesday, 16 June 2004 13:12:14 UTC