>So far I have only grasped the concept that an XML transaction will be an extension of HTML to he point that data items will be "tagged" , where various types of parsing engines will extract the data in a standard way.

I don't know what an XML 'transaction' is, maybe you just chose the wrong words. But you are correct in that various parsers can extract the XML in a standard way. But the more important thing to remember is that there really is no 'standard' way. The way the ML is parsed can be described in the document itself, or in another document referenced by the XML document, called the DTD (Document Type Definition). An XML parser can use the information in the DTD to figure out what to do with the stuff in the document. The DTD, for example might describe a single "cursor" object that has "rows" as children. It might also state that the "row" items each have a set of attributes with names and values. These attributes can be equated to fields. So, using this information, the XML parser can determine that it will be building a table from the information, create the table based on the information, and grab all the infomation in the document and populate the table.


>I imagine therefore that this could be a "flat file" type of thing without relationships etc. etc. and may not be be a data set as we know it.

Yes and no. Mostly no :-). It can be considered a flat file only in that all of the data is stored in one document. But it is not really flat. A single XML document can contain many different data definitions, and even describe how the different types of objects "relate" to each other. This facet of the XML standard allows what equates to a complete relational database to be transported in a single document. Just because the standard allows for realtions though, don't think that relational data is what is limited to. Actually, XML lends itself more naturally to object data design, where the data is laid out hierarchically, children under parent.

All of this flexibility is the true beauty of XML. There really is no limit to the types and formats of the information that can be held in XML.