But now that you have an XML repository, what are you going to do with it? What and how may you deploy it? Simply having large masses of XML converted data doesn’t necessarily mean that the data in this form is even useful.
Norm then explains the problem with using other languages with XML documentbases
The problem with other programming languages isn’t that they aren’t able to process XML, it’s that they aren’t able to process XML efficiently. Data has to be converted from XML to the language’s native data structures. Once converted, it must be manipulated with functions that don’t understand the underlying model and are, consequently, not always a good fit. This “impedance mismatch” causes confusion and can introduce errors. Finally, the programming language structures have to be converted back into XML. Each of these steps is tedious, time consuming, and introduces the possibility of errors. In a sophisticated application, this process may have to occur several times for each XML resource.
If all this looks interesting, the full article is here.