W3C home > Mailing lists > Public > www-ql@w3.org > January to March 2001

Re: [www-ql] <none>

From: Jonathan Robie <Jonathan.Robie@SoftwareAG-USA.com>
Date: Tue, 27 Feb 2001 17:01:01 -0500
Message-Id: <5.0.2.1.0.20010227160856.02b4df98@pop.mindspring.com>
To: Jeff Chapman <Jeff.Chapman@pervasive.com>, www-ql@w3.org
At 03:26 PM 2/26/2001 -0600, Jeff Chapman wrote:

>First and foremost, I really appreciate the work that's being done 
>here.  I really value any effort to improve the ease of XML development 
>and XQuery seems well on it's way to accomplish this.

Thanks - I'm glad you like it.

>In a recent reply to Evan Lenz, Jonathan Robie asserts that "the purpose 
>of FLWR expressions, ... is to provide general SQL-like functionality for 
>joins and declarative restructuring".  Is the charter of the XQuery WG to 
>solve the query problem for XML developers, or to make life easier for SQL 
>applications and SQL programmers.  To me, these seem like two different 
>problem domains.  However, XQuery seems to be more focused on the 
>"SQL-like functionality" issue.

I'm not sure what you mean by "SQL-like functionality". Are you concerned 
about whether we can do queries on structured documents? If so, I think 
this is more a function of the examples in the XQuery document than of 
XQuery itself - we need to improve the set of use cases and examples. XPath 
and XQL, which were a strong influence on XQuery, were designed primarily 
with document scenarios in mind, and you will notice that some of our use 
cases (http://www.w3.org/TR/xmlquery-use-cases) do handle document-oriented 
scenarios like table of contents generation (eg: Use Case "TREE": Queries 
that preserve hierarchy ,  Use Case "SEQ" - Queries based on Sequence, Use 
Case "SGML": Standard Generalized Markup Language, Use Case "TEXT": 
Full-text Search, etc.).

The FLWR expression is indeed SQL-like. Our path expressions are very 
XPath-like, and XPath is probably the most wide used approach to querying 
XML documents.

>IMHO, in our effort to make XQuery look familiar to SQL coders, we're 
>throwing the baby out with the bathwater.  Specifically, I'm deeply 
>troubled that an XQuery expression is not valid XML.  This will 
>immediately cause problems when I try to store XQuery documents into an 
>XML repository.  This should trigger a sense of Deja Vu.  Aren't we 
>ignoring a painful lesson from our immediate past.  When people realized 
>the gravity of the situation caused by DTDs not being valid XML, the W3C 
>immediately began work on a solution--XML Schemas.  Unfortunately, XML 
>Developers are now left scrambling to support two completely different 
>schema definition mechanisms.  If you're an XML tool vendor then this will 
>guarantee that customers will pay for upgrades on a very frequent 
>basis.  However, I would hope we would recognize that this was a mistake 
>and learn from it, instead of repeating that mistake with an XQuery syntax 
>that is not valid XML.

First, our requirements document says that we have to produce an XML 
representation. I would like to see this be semantically identical to the 
XQuery representation, so that one could automatically be converted to the 
other. Using that approach, they would not be two different languages, but 
two different syntaxes for the same language.

>Also, my hope is that we're actually going to solve the key problems which 
>make it really difficult to use XML for data-centric instead of 
>document-centric purposes--and I don't mean to make XML look more like 
>SQL.  Toward that end, the lack of Update capabilities is a huge 
>hole.  While I value the improvements in ease of use, I question the 
>decision to invest in a SQL-like syntax that isn't even valid XML while 
>completely avoiding the critical "Update" issue.  Without this 
>functionality, data repository vendors and application developers will 
>continue to be forced to invested heavily in proprietary tactics that 
>further fragment and isolate XML developers into vendor-specific 
>camps.  Clearly, this is something that the W3C should not encourage, 
>either explicitly or implicitly.

I agree with you, but this is not in our requirements. Several of us have 
been thinking about XQuery updates, but we are in early phases of our 
thinking, and it is not clear to me what relationship this work will have 
to the XML Query activity.

Jonathan

These are my opinions right now. They may be quite different from the 
opinions of Software AG, the W3C XML Query Working Group, or the opinions 
that I will have after reading and considering your response.
Received on Wednesday, 28 February 2001 17:06:08 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Saturday, 22 July 2006 00:10:17 GMT