Re: Minority objection to requiring unique GEDs or required feature to distinguish operations

On Sep 14, 2004, at 11:01 PM, Mark Baker wrote:
> On Tue, Sep 14, 2004 at 01:25:29PM -0400, Amelia A Lewis wrote:
>> On Sep 14, 2004, at 1:16 PM, Mark Baker wrote:
>>> On Tue, Sep 14, 2004 at 01:07:13PM -0400, Amelia A Lewis wrote:
>>>>> But finally, without the requirement, one cannot tell whether two
>>>>> different services with the same binding use the same dispatch
>>>>> mechanism
>>>>> or not, if one is not explicitly mentioned. I'm slightly uneasy 
>>>>> about
>>>>> this.
>>>>
>>>> On the other hand, this gives me warm fuzzies.  The dispatch 
>>>> mechanism
>>>> (or mechanisms) used by the service are probably not properly the
>>>> interest of the users of the service.
>>>
>>> I think this is the crux of the disagreement.  IMO, that information
>>> is critical.  Without it, there is no contract.
>>
>> I couldn't disagree more.
>
> It must be the crux then! 8-)
>
>>  With it, the contract has been fouled with
>> extraneous information that enforces implementation decisions on the
>> service that shouldn't be exposed, much less enforced for the long
>> term.
>
> Hmm, well do you realize that every interface definition language I'm
> familiar with, including OMG/ISO IDL, MIDL, RMI remote interfaces, DCE
> IDL, and every application protocol ever created, has been similarly
> "fouled"?  They each define a dispatch mechanism.  So I think we're on
> pretty safe ground requiring it be unambiguous.

As every IDL that you describe has notable problems, particularly with 
over-tight binding, I think we have identified one of the core problems 
with an "IDL".  WSDL, fortunately, isn't an interface description 
language.  It's a service description language.  Web Services 
Description Language, specifically.

> FWIW though, I was just thinking that if it were specified that an
> agent processing the WSDL should interpret the absence of this
> information as being semantically equivalent to the information being
> unrecognized, then that would be sufficient to address my concerns.
> Does that synch with your view?

Since absence and unrecognized, in the current regime, mean an error, 
then no, I cannot agree that this is any aid.  If the requirement is 
dropped to a recommendation, then it actually becomes more possible to 
say "unrecognized: error; absent: warning" and to trust that the ones 
that are specified actually reflect reality.

While it is a Stupid Requirement, there *will* be WSDL documents that 
contain bogus information.  *shrug*  If it is a recommendation instead, 
then those WSDL that do contain it are *much* more likely to reflect 
reality.

Amy!
-- 
Amelia A. Lewis
Senior Architect
TIBCO/Extensibility, Inc.
alewis@tibco.com

Received on Wednesday, 15 September 2004 15:31:39 UTC