Re: Antw: Re: Sobre las querys en XML

[rodrigo linuxave net]

Vivien Malerba wrote:

> On Fri, Oct 20, 2000 at 10:04:16AM +0200, rodrigo linuxave net wrote:
> > > First some Citations:
> > > 'XQL is a notation for retrieving information from a document'
> > > and :
> > > 'The XQL Proposal, a superset of the XSL pattern syntax, addresses the
> > > information retrieval aspects of queries.'
> > > and:
> > > 'An important area commonly associated with query languages is
> > > database management functions including operations such as
> > > create, delete, update, insert. This area is not currently addressed
> > > by either XML-QL or by  XSL.'
> > >
> >
> > ok, so for the moment, it's not database-oriented but document-oriented?
> >
> > >
> > > Conclusion:
> > > XQL is some (powerfull) kind of (SQL-) select, that allows to extract data
> > > from within a arbitrary complex DOM-tree, but:
> > > - It doesn't support joins
> > > - AFAIK it doesn't support grouping
> > > - it has only one aggregate method count(), that AFAIK does something like
> > >      'SELECTCOUNT(*) FROM table WHERE condition'
> > >
> > > So I think XQL is no substitute for XmlQuery
> > >
> > > Of course there is a non-empty intersection of XQL and XmlQuery:
> > >
> > > If the XmlQuery happens to be a select, that extracts some data from
> > > a single table, it could by replaced by the correnponding XQL equivalent,
> > > because each table could be identified with a simple structured XML document.
> > >
> > > IMO this small intersection does justify a total redesign of XmlQuery.
> >
> > well Gerhard, you and Vivien are the people who had put their work on the
> > XML queries, so I think a redesign of XmlQuery is ok only (and only if) there is
> > something useful to be added to what you've done. So, if you think XML queries are
> > more powerful than this XQL stuff, let's continue like this, and if in the future
> > we find that XQL are useful we can:
> > * add support to XML queries for XQL
> > * add XQL as an option "query language" to the report engine (that is, it will
> > support SQL, XML Queries and XQL)
> >
> > or we could do what Derek says: create our own free standard (just dreaming?)
> >
> > cheers
> >
>
> I also think that it is better to continue with the XML queries. The XQL is
> however an attempt to make queries expressed in a universal way. It is quite
> incomplete, and not really suited to our needs. We can also have a look
> at a specification which is the latest from the W3C regarding XML Query.
>
> at http://www.w3.org/XML/Query there are some interresting things.
> also at http://www.almaden.ibm.com/cs/people/chamberlin/quilt.html there
> is a library which AFAIK implements the XML Query as proposed by the W3C.
> I don't know what its license is though!
>
> As I said I think we should keep with our XML queries model (and thanks to
> Gerhard we will soon have something to test), and keep an eye on the W3C
> XML Query. IF then the XML Query is adopted and IF we need some features
> which are difficult to do with our XML queries THEN we may think about moving
> to the W3C XML Queries. This is my point of view.
>
> There is no point at changing now for no good reason.
>

ok, this is perfect for me. Let's continue with the current XML queries, but, as Vivien
says, keeping an eye on the W3C XML queries, so that in a future, if it's worthy, adapt
our model to that.

cheers