RE: [RTF] Behavior definition of Services - public discussion

Roger

You said ...

>>>I absolutely cannot imagine the people who actually are responsible for
these business processes accepting automated discovery and other automated
things involving semantics other than at the end of a LONG evolutionary
process.<<<

How about this as a use case which describes a "sourcing" activity ...

1. You want a supplier for a commodity product, say 1" no 8 screws.
2. You browse, using your web browser, a directory of product codes to
determine the product code of the product
3. You search a public directory using an XML/SOAP driven search engine for
a supplier who could provide you with the products you need. This returns a
list of suppliers (in an XML document) and also provides details of their
credentials (e.g. Better Business Bureau) to determine that you can
basically trust them
4. You select ten of the suppliers using criteria such as credit, location,
etc, and send them again using XML/SOAP a RequestForQuote (RFQ) asking what
their prices are for delivery of specific quantities of these products at
specific dates/times to specific locations
5. Each supplier has an automated system that provides quotes for the RFQ
which is returned to the buyer
6. You have an automated algorithm that ranks each of the quotes according
to pre-determined criteria
7. The list of quotes is then passed to a user who gets on the phone to the
leading supplier to close the contract and finalize all the other business
terms.

This business process involves:
1. Sending unsolicited messages to both the XML/SOAP driven search engine
and to the individual suppleirs
2. Requires no prior agreement or contract between everyone involved
3. Standardization of the semantics of the messages (i.e. the search query,
the query response, the RFQ and the Quote) and the business process being
followed.
4. Agreed common infrastructure for how the messages are delivered.

I think that Web Servics architecture should be focusing on 4. Other
activities, e.g. UBL are focusing on 3.

Thoughts?

David

-----Original Message-----
From: Cutler, Roger (RogerCutler) [mailto:RogerCutler@ChevronTexaco.com]
Sent: Tuesday, July 09, 2002 2:02 PM
To: 'Francis McCabe'; www-ws-arch@w3.org
Subject: RE: [RTF] Behavior definition of Services - public discussion



I agree rather strongly with your statement, "It is my assertion that that
includes being able to handle the standard (even ancient) ways that business
has been conducted".  Unfortunately, to my mind that means looking at how
EDI works.  REALLY works.  And in general that does NOT include discovery of
business partners or any of this late binding stuff.  The relationships and
processes are set up early and carefully by people.  I absolutely cannot
imagine the people who actually are responsible for these business processes
accepting automated discovery and other automated things involving semantics
other than at the end of a LONG evolutionary process.  The risks would be
too high, and anybody who is doing that sort of business is by nature risk
averse.

In other words, I like your 3 and 4, but probably not 1 (depending on how
one interprets it) and certainly not 2.  And I still agree strongly with
Suresh's proposed statement.

-----Original Message-----
From: Francis McCabe [mailto:fgm@fla.fujitsu.com] 
Sent: Tuesday, July 09, 2002 11:57 AM
To: www-ws-arch@w3.org
Subject: Re: [RTF] Behavior definition of Services - public discussion




On Tuesday, July 9, 2002, at 07:08  AM, Damodaran, Suresh wrote:

>
>
> From the discussions so far in this thread, is there a consensus that 
> "though defining semantic equivalence of functional behavior is an 
> interesting idea, it is very ill defined to be considered a 
> requirement of Web Service Architecture"
>
> Thanks,
>
> -Suresh
> Sterling Commerce

This is a straw man; a very poor implement with which to beat a dead 
horse. On the other hand, simply wishing away a problem doesn't magick 
(sic) it away.

We ignore semantics at our peril. Let me repeat, the game is `doing 
business on the Internet' and `its my system talking to yours' when `we 
met 3 milliseconds ago'. If you can achieve this without explicit 
semantics I'd like to see how to do it.

CORBA `failed' because it required a prior agreement between 
programmers -- i.e., a time scale of months if not years to get stuff 
through OMG. It is also incredibly fragile because of its very early 
binding character. I am not advocating CORBA over SOAP.

On the other hand, `web services as RPC using SOAP' doesn't work either. 
The Internet is a public forum, and that changes everything.

In order to transform web services from `something that shows promise' 
to `something that delivers value to customers' you need to address real 
business needs, not just those that it is easy or convenient to handle. 
It is my assertion that that includes being able to handle the standard 
(even ancient) ways that business has been conducted. It also includes 
taking maximum advantage of the potential offered by the public forum.

If we fail in our vision then we will fail to make a difference and web 
services will be yet another technological roadkill. On the other hand, 
if we adopt a strong goal -- of enabling people to conduct business in a 
standards enabled fashion -- then we will have something to be proud of.

I believe that this calls for:

1. A vision of web services as a deployment platform for doing automated 
business in a public forum
2. A capability for discovering potential business partners
3. A facility of doing business that reflects the natural flow of 
information during a business relationship
4. An environment in which trust and security is fundamental

Frank McCabe



>
>
>
> -----Original Message-----
> From: Newcomer, Eric [mailto:Eric.Newcomer@iona.com]
> Sent: Sunday, July 07, 2002 10:46 AM
> To: Champion, Mike; www-ws-arch@w3.org
> Subject: RE: [RTF] Behavior definition of Services - public discussion
>
>
>
> I think it's best if we concentrate on developing the reference
> architecture
> as "job No. 1" rather than try to reach conclusion on the extent to 
> which
> semantic inferences are integral.
>
> The industry really needs guidance on what a web service is and isn't,
> and
> what is and is not included in a Web services architecture that does 
> more
> than the basics.
>
> Eric
>
> -----Original Message-----
> From: Champion, Mike [mailto:Mike.Champion@softwareag-usa.com]
> Sent: Wednesday, July 03, 2002 4:07 PM
> To: 'www-ws-arch@w3.org'
> Subject: RE: [RTF] Behavior definition of Services - public discussion
>
>
>
>
>
>> -----Original Message-----
>> From: Francis McCabe [mailto:fgm@fla.fujitsu.com]
>> Sent: Wednesday, July 03, 2002 12:18 PM
>> To: Champion, Mike
>> Cc: 'www-ws-arch@w3.org'
>> Subject: Re: [RTF] Behavior definition of Services - public 
>> discussion
>>
>>
>> The bottom line: avoid phrasing the question in terms of equivalence, 
>> instead phrase the question in terms of `have I heard of this name 
>> before'?
>
> My bottom line is
>
>>>> concepts like semantic equivalence that
>>>> could create expectations well beyond what Web Services can 
>>>> actually deliver today.
>
> I'm eagerly looking forward to seeing and using technologies using "a
> graph
> of concepts that a web service provider  publishes to describe his or 
> her
> service. A client applies a matching test to that graph -- which might
> include getting references from other graphs -- to see if the graph is
> congruent with his desired service."  Maybe I'm not looking in the right
> places, but I just don't see that in the real world of web services 
> today.
>
> Thus, it is IMHO inappropriate to *require* the WSA to accomodate 
> ideas which *may* prove powerful, until their practical value has been 
> demonstrated.  The W3C -- to bang one of my favorite drums, sorry -- 
> is most successful when working to standardize practice, and least 
> successful when
> trying to do computer science by committee.  I would be very happy to
> incorporate field-tested semantic inference technology into the WSA, 
> but I
> can't agree to require it based on the current state of the art.
>