W3C home > Mailing lists > Public > www-ws-desc@w3.org > July 2006

RE: WSDL Import with just a location attribute

From: Jonathan Marsh <jmarsh@microsoft.com>
Date: Thu, 13 Jul 2006 19:43:26 -0700
Message-ID: <37D0366A39A9044286B2783EB4C3C4E8034C8C2F@RED-MSG-10.redmond.corp.microsoft.com>
To: "Ramkumar Menon" <ramkumar.menon@gmail.com>, "Arthur Ryman" <ryman@ca.ibm.com>
Cc: <www-ws-desc@w3.org>, <www-ws-desc-request@w3.org>
I don't think it makes it easier to discover the "actual documents
imported", by which I assume you mean the set of components defined in a
particular namespace.  It makes discovery of these components more
flexible, as an implementation can provide a variety of ways to find
those components besides just parsing the document referenced by the
location attribute.  It might have built-in copies.  It might cache
copies.  It might have a pre-populated library.


In the simplest case, it seems to me that an implementation that always
relies on the location attribute is effectively a "document" import.
That is a straightforward implementation strategy, but might not offer
as great a potential for optimization as keying imports off the
namespace (instead of the location) does.



From: www-ws-desc-request@w3.org [mailto:www-ws-desc-request@w3.org] On
Behalf Of Ramkumar Menon
Sent: Thursday, July 13, 2006 6:49 PM
To: Arthur Ryman
Cc: www-ws-desc@w3.org; www-ws-desc-request@w3.org
Subject: Re: WSDL Import with just a location attribute


Hi Arthur, 


Thanks a lot for your response.

Actually, I am in absolute congruence with you on the current semantics
of import. 

My question [might be extremely silly, but I am just trying to convince
myself :-( ] is on the lines of 

What is the rationale behind choosing to use <import> as a declaration
thats used only for importing namespaces ? Why cant it be on documents
as well ? Doesnt import of a document correspond indirectly to the
import of the tasrget namespace of the document - or rather why cant we
assume that intent ? 


b) Is it true that in majority of the use-cases, 

  i) It makes the job easier for the processing application to define
its own way of discovering the actual documents imported into the main
document corresponding to the namespaces defined in the import ? or
rather, is it the case that in majority of the use-cases, the processing
application does not use the location attribute of the <import> ? 

 ii) Are the imported documents usually not known at design time ? If
they can be discovered at designtime, wdnt it be an easier task to
specify the location within the wsdl directly ?




On 7/13/06, Arthur Ryman <ryman@ca.ibm.com> wrote: 


The <import> element is a declaration that the document references
components from another namespace. It imports the namespace, not a
document. The location attribute is just a hint of where to find the
definitions of the components in the namespace. 

If you just want in bring in the contents of a document, use <include>
but that requires that the included document have the same namespace. 

We decided to make targetNamespace required to simplify the spec and
avoid problems found with WSDL 1.1 where targetnamespace was optional.
For example, targetname typically maps to a package name when generating
code and the way "default" packages are handled by various tools is

Arthur Ryman,
IBM Software Group, Rational Division

blog: http://ryman.eclipsedevelopersjournal.com/
phone: +1-905-413-3077, TL 969-3077
assistant: +1-905-413-2411, TL 969-2411
fax: +1-905-413-4920, TL 969-4920
mobile: +1-416-939-5063, text: 4169395063@fido.ca 

"Ramkumar Menon" < ramkumar.menon@gmail.com
<mailto:ramkumar.menon@gmail.com> > 
Sent by: www-ws-desc-request@w3.org 

07/12/2006 06:49 PM 






WSDL Import with just a location attribute








Hi Gurus,
What is fundamentally wrong with an import of a document, rather than
a foreign namespace, into a WSDL Document ?
If I translate my thoughts into syntax, 
why is
<wsdl:import location=" http://www.twinktwinklilstar.com/wsdls/star.wsdl
<http://www.twinktwinklilstar.com/wsdls/star.wsdl> "/>
so unholy ?

Why not accept this variant ["of course", in addition to the existing
version] as a valid wsdl import ? If not, what's the strong reason 
against this?

One more question - why is targetNamespace attribute for WSDL 2.0
documents mandatory ? Is it a WS-I issue ?

Shift to the left, shift to the right!
Pop up, push down, byte, byte, byte! 

-Ramkumar Menon
A typical Macroprocessor

Shift to the left, shift to the right!
Pop up, push down, byte, byte, byte!

-Ramkumar Menon
A typical Macroprocessor 
Received on Friday, 14 July 2006 02:44:23 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 23:07:00 UTC