Cart
[Home] [By Thread] [By Date] [Recent Entries]
Dear all I brought the Update issue up, and I would like to refocus the discussion. I see that XPath, XSLT and XQuery is all very much designed to work with text retrieval and formatting. Maybe you could call these techniques "client oriented." I see that e.g. SQL is quite different, since it is designed to work with data: Definition, Retrieval, Manipulation. The funny issue is that the XML brigade wants to move from the original document/specialist enduser oriented SGML world, into the huge lucrative market of DBMS. And here it is not enough with some nice client oriented retrieval and formating languages, but a fully built real Data Definition (DDL) and Data Manipulation language (DML). And to become this, XQuery must incoorperate Update, as well as the DBMS thinking in the design, instead of just retrieval of information thinking. I have had the joy of seing Jonathan demonstrate XQuery, and it looks good on paper for XML retrieval and formating (In fact, allmost all of the other people in the audience said: "Wow. We will definately not buy any XML DB without XQuery support...." after having seen XQuery vs. Xpath) So in short: If XQuery goes ahead without Update, it will become a nice tool, but not a real usefull standard. And we will see independent update semantics emerge from Oracle, IBM, MS etc. and chaos will prevail. I still think that XQuery should include Update. Else we will have to wait another 4 years for W3C standard on this area, and it will be too late. JJ
|
