Re: Etymology of "web service"

I think Andrew has it exactly right.  I would point especially to:  "Also
the observation that the result could be a web of
dynamically interacting services.", and I would add the crucial fact that
web services are identified with URI's.  That's a key step in making these
services part of "the" web rather than "a" web.  While a lot of the
infrastructure to do this is still being built, and while many services
will be strictly machine to machine, it is a very cool thing that you can
in principle put a link to a web service on your home page.  One day, your
browser may even know how to do something smart with the SOAP message that
comes back when you click the link.

Note that XML Schema and SOAP Encoding give you the xsd:anyURI type, which
allows you to indicate that the value returned from a SOAP call is itself
a URI, and that URI might well be the handle on yet another web service,
possibly one created by the first call.  So, if I use SOAP to open a bank
account, I can get back the URI of a service that represents just my
account.  I fervently hope that WSDL will evolve to support this idiom in
a convenient and robust manner.

The fact that we are talking about a web of services captures part of the
reason that Web Services are interestingly different from COM and Corba.
We're talking about loose coupling, late binding, and dynamic location of
services.  These are things that the WWW already does on a global scale.
Note: that the so-called "REST" features that were recently added to SOAP
1.2 (e.g. the ability to do an HTTP GET, when you're retrieving as opposed
to updating information) are very useful in leveraging the power of "the
Web" to build web services.  See, for example [1,2].  The HTTP binding in
SOAP 1.2 supports GET as well as POST [3].

Noah

[1] http://www.w3.org/TR/2002/WD-soap12-part2-20020626/#WebMethodFeature
[2] http://www.w3.org/TR/2002/WD-soap12-part2-20020626/#RPConWeb
[3] http://www.w3.org/TR/2002/WD-soap12-part2-20020626/#soapinhttp

------------------------------------------------------------------
Noah Mendelsohn                              Voice: 1-617-693-4036
IBM Corporation                                Fax: 1-617-693-8676
One Rogers Street
Cambridge, MA 02142
------------------------------------------------------------------







Andrew Layman <[email protected]>
Sent by: Discussion of implementing SOAP applications today
<[email protected]>
11/06/2002 02:43 PM
Please respond to Discussion of implementing SOAP applications today


        To:     [email protected]
        cc:     (bcc: Noah Mendelsohn/Cambridge/IBM)
        Subject:        Re: [SOAP] Etymology of "web service"


I think the Web part comes from the observation that the technologies
used underneath Web services are the same technologies that made the WWW
possible: loosely-speaking DNS, TCP/IP, HTTP, SMTP and XML (which is
similar to HTML but more suited for extensible, structured data
exchange).  Also the observation that the result could be a web of
dynamically interacting services.

The name has caused or permitted a fair amount of confusion, because,
while the name captures some of the similarities to the WWW, it also can
be used to make sound plausible some pretty off-base claims such as
"Broadly defined, a Web service is any service delivered across the
Web".

-----Original Message-----
From: Chuck Hinson [mailto:[email protected]]
Sent: Wednesday, November 06, 2002 10:27 AM
To: [email protected]
Subject: Re: [SOAP] Etymology of "web service"

Andrew Layman wrote:

>Etymology won't tell you what a web service is.  The concept is not a
>simple consequence of the two words.  Definition: A web service is an
>interoperable computational service, specified in terms of protocols
>(the messages it receives and sends), language-neutral,
>platform-independent, using XML and SOAP, and based on
>computer-to-computer message exchange without the assumption of a human
>as the immediate processor of the messages.
>

It's interesting that the definition says nothing about the web.  I've
always wondered why they're called *Web* services.

--Chuck

>
>
>-----Original Message-----
>From: Sam [mailto:[email protected]]
>Sent: Wednesday, November 06, 2002 6:47 AM
>To: [email protected]
>Subject: [SOAP] Etymology of "web service"
>
>I was curious to see if anyone knew the origin of the term
>"web service" . Trying to see if I can get the etymology behind it.
>
>Thanks
>/s
>
>You can read messages from the SOAP archive, unsubscribe from SOAP, or
>subscribe to other
>DevelopMentor lists at http://discuss.develop.com.
>
>You can read messages from the SOAP archive, unsubscribe from SOAP, or
subscribe to other
>DevelopMentor lists at http://discuss.develop.com.
>

You can read messages from the SOAP archive, unsubscribe from SOAP, or
subscribe to other
DevelopMentor lists at http://discuss.develop.com.

You can read messages from the SOAP archive, unsubscribe from SOAP, or
subscribe to other
DevelopMentor lists at http://discuss.develop.com.

You can read messages from the SOAP archive, unsubscribe from SOAP, or subscribe to other
DevelopMentor lists at http://discuss.develop.com.