OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help



   Re: [xml-dev] Yet another plea for XUpdate (was New XSLT, XPath, and XQu

[ Lists Home | Date Index | Thread Index ]

> At 01:17 PM 5/2/2002 -0600, Uche Ogbuji wrote:
> > > I know it's a near hopeless case, politically, but I
> > > have no idea why there is not more consideration of
> > > XUpdate in these discussions.
> [ !!! SNIP !!! ]
> >http://www.xmldatabases.org/projects/XUpdate-UseCases/
> > > http://www.xmldb.org/xupdate/
> > >
> > > One problem, for sure, is that the spec needs
> > > polish.
> Depends what you mean by "these discussions". In the XML Query Working 
> Group,

This is not what I mean at all.  I hardly think of XQuery as the only game in town for querying XML docs and doc collections.  No more so than I think W3C XSDL is the only game in town for XML schema definition.

The particular sentiment I was responding to concerned Vendors each going their own way in the implementation of proprietary update systems.

> I think people want an update syntax that is a natural extension of 
> the existing XQuery language, a design goal which XUpdate does not try to 
> address.

No offence whatsoever to you or the XML Query group, but I don't expect to be using XQuery very much, so XQuery integration is the least of my concerns for an XML update language.

> XUpdate uses an XML vocabulary to express where updates should be made, eg:
> <xupdate:modifications version="1.0"
> xmlns:xupdate="http://www.xmldb.org/xupdate";>
> <xupdate:insert-before select="/addresses/address[@id = 1]/name/last" >
> <xupdate:element name="middle">Lennox</xupdate:element>
> </xupdate:insert-before>
> </xupdate:modifications>
> In the update proposal I presented at XML 2001, which is virtually 
> identical to the proposal in the Lehti paper, the equivalent query would be:
> update
>    insert <middle>Lennox</middle>
>    preceding document("xupdate.xml")/addresses/address[@id = 1]/name/last

I don't find this verbosity any more troubling than the same property in XSLT.  Verbosity is not much of a consideration for me in XML Update.

I'll tell you one reason why I prefer XUpdate's insiatence on XML syntax: it makes implementation a breeze.  Seriously.  4Suite's XUpdate is entirely implemented in 500 lines of Python and took about a day to complete.  There is *no* way on earth that I could implement the second syntax you put forth so easily.

Dare's syntax looks just as hairy to implement.

> Integration with the rest of XQuery gives you one other important 
> advantage: type safety. That really matters for updates.

Well, as you know, I think the conceit of "type safety", as it is conventionally used in W3C XML specifications, is lifted without much thought from the blinkered models of typing in relational and object theory.  This makes it, IMO, a titanic waste of productivity.

I certainly don't see how typing matters so much for updates.  What matters is schema consistency, which may or may not involve data typing considerations.

Uche Ogbuji                           Fourthought, Inc.
uche.ogbuji@fourthought.com           http://fourthought.com
http://4Suite.org                     http://uche.ogbuji.net
Track chair, XML/Web Services One (San Jose, Boston): http://www.xmlconference.com/
RDF Query using Versa - http://www-106.ibm.com/developerworks/xml/library/x-think10/index.html
WSDL and the Wild, Wild West - http://adtmag.com/article.asp?id=6004
XML, The Model Driven Architecture, and RDF @ XML Europe - http://www.xmleurope.com/2002/kttrack.asp#themodel


News | XML in Industry | Calendar | XML Registry
Marketplace | Resources | MyXML.org | Sponsors | Privacy Statement

Copyright 2001 XML.org. This site is hosted by OASIS