Posted 19 May 2006 - 03:06 PM
XML is so much more than just delivering up web pages.... an XML file is a self contained database of sorts. HTML is used to tell how to display some content. XML gives information about the structure and symantics of that content. unlike HTML, XML users can define their own data tags. If you wrap XML in RDF, then not only are you able to tell information about structure and syntax, but you are able to identify relations between data objects. One of the cool things about XML is that you can pull xml from a variety of sites and combine them in such a way to create an entirely new datatype. For instance, if you wanted to research how weather effects bird migratory patterns, you could combine data from any weather service, bird species, and migration patterns, and combine them in such a way to find correlations between the data... I personally find that very cool.
XML is also useful for web applications, because only the data that has changed needs to be transmitted back and forth, thus cutting down on the need to hit the database across the network (although the trade off is sometimes a longer initial load time).
Different XML files can be parsed and combined to create brand new XML files. By using schemas or DTDs, you can verify that the integrity of the XML file, and verify the correctness of your data. Granted, XML is somewhat limited because it can not define data restraints.
..and it can deliver web pages, or even RSS.
SQL is a language for quering and manipulating data in relational databases, and doesn't necessarily have anything to do with XML, although XML data can be stored in a relational database. There are many other ways to store data.