W3C home > Mailing lists > Public > www-tag@w3.org > October 2002

RE: Potential TAG issue in re consistency, Schema, etc.

From: Dare Obasanjo <dareo@microsoft.com>
Date: Sun, 13 Oct 2002 22:35:36 -0700
Message-ID: <B885BEDCB3664E4AB1C72F1D85CB29F8040EF649@RED-MSG-10.redmond.corp.microsoft.com>
To: "Jeni Tennison" <jeni@jenitennison.com>, <www-tag@w3.org>, "Tim Bray" <tbray@textuality.com>

Disclaimer: The following comments reflect my opinion and do not represent the my employer in any way, shape or form. 
 
XPath 2.0 and XQuery use W3C XML Schema as a basis for a type system and not merely as a validation language. There is currently no XML schema language I have come across that can satisfy this need for both languages.  Instead of dwelling in hypotheticals it would progress this discourse if you presented alternatives instead of simply stating your dislike for the current state of affairs. "W3C XML Schema sucks and XPath shouldn't use it" is an opinion and not a Web architectural issue. 

	-----Original Message----- 
	From: Jeni Tennison [mailto:jeni@jenitennison.com] 
	Sent: Sat 10/12/2002 11:08 AM 
	To: www-tag@w3.org; Tim Bray 
	Cc: 
	Subject: Re: Potential TAG issue in re consistency, Schema, etc.
	
	


	Hi Tim,
	
	> The XQuery Working Group's response continues "While this is clearly
	> a decision for the Architecture team..." which I guess means the
	> TAG.
	>
	> I'm hardly unbiased here because I strongly disagree with the
	> direction the XQuery WG is going here on technical grounds.
	> Furthermore, I think that if this issue were being discussed on
	> purely technical terms, some decisions might be different.
	>
	> So does anyone think there's an issue here we ought to take up?
	
	Yes, please. I'll note, though, that this is also an XPath issue.
	Personally, I'm concerned that the XQuery design decisions are having
	a large effect on the design of XPath 2.0, which of course has an
	effect on XSLT 2.0. You've probably noticed the discussions on XML-Dev
	in which users and vendors state that they won't be using/implementing
	XSLT 2.0 because of the burden of XPath 2.0. I think that this effect
	on two of the core, and more successful, XML-related technologies is
	very worrisome.
	
	There are several issues in the XPath/XQuery space that I think count
	as architectural-level issues:
	
	  1. The relationship between XPath and XML Schema Structures --
	     whether XML Schema 1.0 is the only schema language that XPath
	     should cater for or whether it should be extensible such that
	     other schema languages can be "plugged in" now or at a later
	     date.
	
	  2. The relationship between XPath and W3C XML Schema Datatypes --
	     the extent to which XPath should adopt the data types specified
	     by XML Schema 1.0 and if so whether making changes/additions to
	     this set is appropriate.
	    
	  3. The relationship between XPath and XQuery -- whether XQuery
	     should, as it does currently, "subsume" XPath, or whether XPath
	     should be designed in such a way that XQuery can act as a host
	     language for it, like any other.
	
	  4. The relationship between XPath and other standards that use it,
	     such as XPointer, XForms, DOM Level 3 (XPath), even the XML
	     representation of XML Schema -- whether these should be
	     considered host languages for XPath 2.0 and, if so, whether XPath
	     2.0 meets their requirements.
	
	  5. The relationship between XPath 2.0 and XPath 1.0 -- whether XPath
	     2.0 should be the natural successor to XPath 1.0, to which all
	     host languages are requited to upgrade, or should be viewed as an
	     enhanced technology that is only relevant to technologies
	     requiring schema support.
	
	If these *are* within the TAG's remit (and I think they are), I hope
	that you'll consider them.
	    
	Cheers,
	
	Jeni
	
	---
	Jeni Tennison
	http://www.jenitennison.com/
	
	
Received on Monday, 14 October 2002 01:36:12 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 12:47:12 GMT