W3C home > Mailing lists > Public > xml-dist-app@w3.org > January 2002

Re: Repost issue: missing accessors same as NILs?

From: Jacek Kopecky <jacek@systinet.com>
Date: Fri, 18 Jan 2002 18:53:37 +0100 (CET)
To: Rich Salz <rsalz@zolera.com>
cc: <xml-dist-app@w3.org>
Message-ID: <Pine.LNX.4.33.0201181837240.32025-100000@mail.idoox.com>
 Rich, 
 the first question is really an implementation choice of the
particular soapbuilder and his SOAP stack.
 Regarding your "can I rely" question: How can you even rely that
on the other side there will be a processor that processes your
echoString RPC call in the right way? Or, if it does return to
you the right results, how can you be sure that it didn't in fact
do something completely different and was only masking that by
sending you seemingly correct results?
 The completely satisfactory answer to the last question and IMO
to the questions in etc. is "the application must specify that."  
And that is not the application implemented on the server, it's
the application's desciption around which both the client and the
server (if there are such roles in this application) are built. 
It's the contract.

                   Jacek Kopecky

                   Senior Architect, Systinet (formerly Idoox)
                   http://www.systinet.com/



On Fri, 18 Jan 2002, Rich Salz wrote:

 > >  IMHO "no value is known" is covered by "default" because the 
 > > application can choose to treat missing data as unknown if it 
 > > can choose the default.
 > 
 > 
 > I believe the difference between default and unspecified is worth 
 > calling out.  Default raises some questions -- does the application have 
 > to specify a default value out of band somehow?  Can I rely on a default 
 > value at the other side?  Must all omitted elements of the same data 
 > object have the same default?  Etc.
 > 	/r$
 > 
Received on Friday, 18 January 2002 12:53:39 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:59:06 GMT