W3C home > Mailing lists > Public > www-ws-desc@w3.org > September 2003

RE: "Bulk load" get/set

From: Tom Jordahl <tomj@macromedia.com>
Date: Tue, 30 Sep 2003 10:37:41 -0400
Message-ID: <CB1FF0A474AEA84EA0206D5B05F6A4CB048DF954@S1001EXM02.macromedia.com>
To: "'www-ws-desc@w3.org'" <www-ws-desc@w3.org>


There was a more complicated proposal that was presented in the last F2F meeting in North Carolina that included the bulk operations.

This proposal was deemed way to complex and the proposal made in Palo Alto was the result of some good work to limit the scope of including attributes in to WSDL.

I will go on record as saying that while attributes seem like a useful concept, I do no think they need to be part of WSDL.  They can easily be modeled in an interface, and the extension mechanism (aka interface inheritance) we now have makes it very simple to define a set of attributes that apply to a particular interface.

--
Tom Jordahl
Macromedia Server Development

-----Original Message-----
From: Mark Baker [mailto:distobj@acm.org] 
Sent: Tuesday, September 30, 2003 8:57 AM
To: Jim Webber
Cc: www-ws-desc@w3.org
Subject: Re: "Bulk load" get/set


On Tue, Sep 30, 2003 at 03:46:47AM -0400, Jim Webber wrote:
> Hey Mark, 
> 
> > So, I propose that interfaces with attributes implicitly 
> > support two operations, GET and PUT.  GET would return an XML 
> > document which contains a serialized form of all attributes, 
> > and PUT would accept an XML document and update the value of 
> > the attributes to those reflected in that document.
> 
> This takes WSDL from being a language for building interfaces (and you know
> that I don't really mean "interfaces" in the classical sense there), into an
> interface.
>
> Surely if attributes are required (and that case has really only been
> supported in strength by the Gridf community), then how to expose those
> attributes is an application-level thing?

Erm, did I misread the minutes?  It seems to me that the WG *already*
decided that get/set operations would be used.  That doesn't prescribe
a binding, of course, but neither does my proposal; that the operations
I proposed are "GET" and "PUT" doesn't prevent a binding to FTP RETR
and STOR, for example.

Mark.
-- 
Mark Baker.   Ottawa, Ontario, CANADA.        http://www.markbaker.ca
Received on Tuesday, 30 September 2003 10:37:35 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:58:26 GMT