RE: Words for the Triangles

reply:

I agree with all of David's points regarding Google and other similar search
engines.  If Web Services were visible from a Google, there is no guarantee
that all of the requisite devTime or runTime WbS information would be
present within the Google Results set.  This would depend upon the crawler
that made the information available to the Google "aggregators."

I think what distinguishes a usable registry from the "web" in david's reply
below is that the standards-based information that is available from within
a web services registry (based on UDDI) forms the basis for its uses.  In
other words, I can know how to use the web services metadata information
because it is provided in a standards-based way and its uses and patterns
are predictable.

Joel

-----Original Message-----
From: David Booth [mailto:dbooth@w3.org]
Sent: Monday, September 30, 2002 4:27 PM
To: Heather Kreger
Cc: www-ws-arch@w3.org
Subject: Re: Words for the Triangles




Heather Kreger <kreger@us.ibm.com> on Thu, 26 Sep 2002 17:54:16 -0400
writes:
>No, you do not discover from the web. Google discovers from the Web. Then
>he records it in his registry. THEN you discover from Google,  which is
>searching his - gasp - registry. So, you are discovering from a registry. 
>. . .

Well, yes, that is a legitimate one way to look at it, but I certainly do 
not think of it that way.

To me, Google and other tools are simply my MEANS of discovery -- not the 
desired end point of my discovery.  Google, Yahoo, email, and postcards and 
billboards with URIs on them are all means that I use to find what's on the 
Web.  Google just helps me to find it by suggesting a URL, but I get the 
information that I seek from the Web -- I don't get it from Google.  All I 
get from Google are suggestions for where I should look.

To me, the top cloud in the triangle diagram is simply the Web -- that 
universal space where you can put information and get information.  And a 
UDDI registry is certainly a part of that space.  A Web Service publicizes 
its information (somehow), and a Client finds that information 
(somehow).  But they don't have to use a common "registry" to do so (unless 
you wish to call the entire Web a "registry", but I think that would be 
rather disingenuous).  The only thing they have to have in common is the 
Web: One places the information on the Web and the other finds it.


-- 
David Booth
W3C Fellow / Hewlett-Packard
Telephone: +1.617.253.1273

Received on Wednesday, 2 October 2002 15:18:33 UTC