W3C home > Mailing lists > Public > www-ws-desc@w3.org > February 2004

RE: Second level xs:import

From: Martin Gudgin <mgudgin@microsoft.com>
Date: Thu, 5 Feb 2004 01:39:49 -0800
Message-ID: <DD35CC66F54D8248B6E04232892B633801188FCE@RED-MSG-43.redmond.corp.microsoft.com>
To: <ygoland@bea.com>, "Prasad Yendluri" <pyendluri@webmethods.com>, <www-ws-desc@w3.org>
IMO the text "Using the import construct is a necessary condition for
making components from another namespace available to a WSDL
description." is accurate language describing the behaviour. I don't see
why any further text is needed in the core spec.
 
Gudge


________________________________

	From: www-ws-desc-request@w3.org
[mailto:www-ws-desc-request@w3.org] On Behalf Of Yaron Goland
	Sent: 05 February 2004 01:33
	To: 'Prasad Yendluri'; www-ws-desc@w3.org
	Subject: RE: Second level xs:import
	
	
	I must admit that I'm confused about the rules used to decide
what goes into the core spec and what goes into the primer.
	 
	My understanding is that an implementer should never need to
read anything but the core specs in order to implement. A primer, to my
understanding, is the moral equivalent of an O'Reilly book. Nice to have
but it's not the reference and references need to be self-contained.
This includes explanations for any ambiguities or non-intuitive
behaviors.
	 
	The current WSDL spec contains explanatory non-normative text of
the exact same nature as what I propose below. 
	 
	For example, section 4.2 has the follow explanatory paragraph "
Specifically it can be used to import components from WSDL descriptions
that do not share a target namespace with the importing document.
Components in imported descriptions are part of the component model of
the importing description. The imported components can be referenced by
QName. Using the import construct is a necessary condition for making
components from another namespace available to a WSDL description. That
is, a WSDL description cannot refer to components in a namespace other
that the target namespace unless an import statement for that namespace
is present. "
	 
	If the previous paragraph is appropriate for the core spec and
not for the primer then why is the text I proposed inappropriate for the
core spec?
	 
	            Just Curious,
	 
	                    Yaron

		-----Original Message-----
		From: www-ws-desc-request@w3.org
[mailto:www-ws-desc-request@w3.org]On Behalf Of Prasad Yendluri
		Sent: Wednesday, February 04, 2004 5:16 PM
		To: www-ws-desc@w3.org
		Subject: Re: Second level xs:import
		
		
		Putting it in the primer makes sense to me as well.
Thanks.
		
		-------- Original Message -------- 
Subject: 	Re: Second level xs:import	
Date: 	Thu, 5 Feb 2004 06:15:32 +0600	
From: 	Sanjiva Weerawarana <sanjiva@watson.ibm.com>
<mailto:sanjiva@watson.ibm.com> 	
To: 	Prasad Yendluri <pyendluri@webmethods.com>
<mailto:pyendluri@webmethods.com> , <www-ws-desc@w3.org>
<mailto:www-ws-desc@w3.org> 	
References: 	<074601c3eb68$bfc1c860$65e5e40c@bea.com>
<mailto:074601c3eb68$bfc1c860$65e5e40c@bea.com>
<40216C5B.7020908@webmethods.com>
<mailto:40216C5B.7020908@webmethods.com> 	

		+1 for putting in the primer. This does not belong in
the core spec.
		
		In any case, if we do our jobs right about 10 people
will read the
		core spec and millions (actually probably billions ;-))
will read the
		primer.
		
		Sanjiva.
		
		----- Original Message -----
		From: "Prasad Yendluri" <pyendluri@webmethods.com>
<mailto:pyendluri@webmethods.com> 
		To: <www-ws-desc@w3.org> <mailto:www-ws-desc@w3.org> 
		Sent: Thursday, February 05, 2004 4:04 AM
		Subject: Re: Second level xs:import
		
		
		> I am in favor of clarifying this via some explanatory
text in the spec.
		> This is a typical confusion and the members of this WG
itself have gone
		> around this topic several times already, each time
resulting in long
		> winded threads. I am sure many WSDL users that are
novice and other that
		> are though beyond novice,  yet to achieve expert level
:) in schema
		> usage would find this helpful. When this issue shows
up again we can
		> simply point to the right spot in the spec or say
RTFS.
		>
		> Prasad
		>
		> -------- Original Message --------
		> Subject: RE: Second level xs:import
		> Resent-Date: Wed, 4 Feb 2004 16:49:31 -0500 (EST)
		> Resent-From: www-ws-desc@w3.org
		> Date: Wed, 4 Feb 2004 13:49:21 -0800
		> From: Yaron Goland <ygoland@bea.com>
<mailto:ygoland@bea.com> 
		> Reply-To: <ygoland@bea.com> <mailto:ygoland@bea.com> 
		> To: 'Amelia A Lewis' <alewis@tibco.com>
<mailto:alewis@tibco.com> , 'Roberto Chinnici'
		> <Roberto.Chinnici@Sun.COM>
<mailto:Roberto.Chinnici@Sun.COM> 
		> CC: <www-ws-desc@w3.org> <mailto:www-ws-desc@w3.org> 
		>
		> While I can understand the motivation I would submit
that the behavior is
		> not intuitive. At a minimum it would seem appropriate
to include an
		> explanation of the behavior so that users and
implementers will understand
		> what is going on.
		>
		> I therefore move that the following text be added to
the end of the first
		> paragraph in section 3.1.1 as defined in the
2004/01/06 22:48:24 draft:
		The
		> prohibition against directly referring to components
imported by schemas
		> that are themselves imported/included by WSDL is
intended to force WSDL
		> definitions to directly refer to the schema
definitions of all components
		> used by that WSDL. Requiring direct references to
imported schema
		components
		> is felt to provide a cleaner WSDL definition. While
this restriction may
		> make it look as if the same schema is being imported
multiple times it is
		> not actually so. For example, Schema A imports Schema
B which has
		components
		> needed by Schema A's definitions. A WSDL that imported
Schema A is free to
		> reference any component in Schema A but the components
in Schema B are,
		> effectively speaking, invisible to the WSDL. While
Schema A can refer to
		> Schema B components, the WSDL cannot. If the WSDL
needs to directly
		> reference schema B components then the WSDL must
explicitly import Schema
		B.
		> Only when Schema B is explicitly imported by WSDL will
Schema B's
		components
		> become visible to the WSDL.
		>
		> Yaron
		>
		> > -----Original Message-----
		> > From: www-ws-desc-request@w3.org
[mailto:www-ws-desc-request@w3.org]On
		> > Behalf Of Amelia A Lewis
		> > Sent: Wednesday, February 04, 2004 7:52 AM
		> > To: Roberto Chinnici
		> > Cc: www-ws-desc@w3.org
		> > Subject: Re: Second level xs:import
		> >
		> >
		> >
		> > Good analogy.  +1.
		> >
		> > On Tue, 03 Feb 2004 17:34:57 -0800
		> > Roberto Chinnici <Roberto.Chinnici@Sun.COM>
<mailto:Roberto.Chinnici@Sun.COM>  wrote:
		> >
		> > > It doesn't have anything to do with inheritance,
it's
		> > really all about
		> > > modules.
		> > >
		> > > An xsd:import means "I'm going to use some
top-level components from
		> > > this namespace, please make them available".
Perhaps those
		> > components
		> > > need in turn to use components from yet another
namespace, but why
		> > > should I see them? They are an implementation
detail really.
		> > >
		> > > Or, to use a programming language analogy,
xsd:import and
		> > wsdl:import
		> > > are more like Java import, not C #include. And
that's the correct
		> > > definition, IMHO.
		> > >
		> > > Roberto
		> > >
		> > >
		> > > Yaron Goland wrote:
		> > > > If I import a Schema file from namespace Foo and
the Schema File I
		> > > > imported itself imports a schema file from
namespace Bar then
		> > > > effectively the WSDL file has imported namespace
Bar as well and
		> > > > should be free to reference Bar. The inheritance
chain is
		> > clear. The
		> > > > namespaces are all explicitly declared. What's
the problem?
		> > > >
		> > > >
		> > > >>-----Original Message-----
		> > > >>From: Martin Gudgin
[mailto:mgudgin@microsoft.com]
		> > > >>Sent: Friday, January 30, 2004 8:18 AM
		> > > >>To: ygoland@bea.com; Amelia A Lewis; David
Orchard
		> > > >>Cc: www-ws-desc@w3.org
		> > > >>Subject: RE: Second level xs:import
		> > > >>
		> > > >>
		> > > >>The *design* limitation, was that schema wanted
people to be
		> > > >>*explicit*
		> > > >>about namespaces they wanted to use. So, in
order to reference
		> > > >>components in namespace foo, a schema MUST have
an import for
		> > > >>namespace
		> > > >>foo ( or itself be a schema for namespace foo ).
		> > > >>
		> > > >>I think it is a reasonable design decision to
make for WSDL too.
		> > > >>
		> > > >>Gudge
		> > > >>
		> > > >>
		> > > >>>-----Original Message-----
		> > > >>>From: www-ws-desc-request@w3.org
		> > > >>>[mailto:www-ws-desc-request@w3.org] On Behalf
Of Yaron Goland
		> > > >>>Sent: 26 January 2004 17:30
		> > > >>>To: 'Amelia A Lewis'; 'David Orchard'
		> > > >>>Cc: www-ws-desc@w3.org
		> > > >>>Subject: RE: Second level xs:import
		> > > >>>
		> > > >>>
		> > > >>>While I can appreciate the wisdom in re-use,
re-use should
		> > > >>>only be done with open eyes and full
understanding. Do we
		> > > >>>know the technical reason why the restriction
is there? If
		> > > >>>not then we should either find out or remove
the restriction.
		> > > >>> Thanks,
		> > > >>> Yaron
		> > > >>>
		> > > >>>
		> > > >>>>-----Original Message-----
		> > > >>>>From: www-ws-desc-request@w3.org
		> > > >>>
		> > > >>>[mailto:www-ws-desc-request@w3.org]On
		> > > >>>
		> > > >>>>Behalf Of Amelia A Lewis
		> > > >>>>Sent: Monday, January 26, 2004 12:05 PM
		> > > >>>>To: David Orchard
		> > > >>>>Cc: www-ws-desc@w3.org
		> > > >>>>Subject: Re: Second level xs:import
		> > > >>>>
		> > > >>>>
		> > > >>>>
		> > > >>>>Because that works the same way that schema
import does,
		> > > >>
		> > > >>and that's what it's modeled on.
		> > > >>>>
		> > > >>>>Amy!
		> > > >>>>On Jan 26, 2004, at 2:54 PM, David Orchard
wrote:
		> > > >>>>
		> > > >>>>
		> > > >>>>>Why is it illegal to reference items that are
included in an
		> > > >>>>>imported/included schema vis xs:import? (per
section 3 of part 1)
		> > > >>
		> > > >>>>>Cheers,
		> > > >>>>>Dave
		> > >
		> >
		> >
		> > --
		> > Amelia A. Lewis
		> > Architect, TIBCO/Extensibility, Inc.
		> > alewis@tibco.com
		> 

				    
Received on Thursday, 5 February 2004 04:40:08 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 5 February 2014 07:15:02 UTC