Microsoft sure adds a lot of extra (unnecessary?) data inside the xml version of those documents. My test ".doc" in its standard document format is only 24k in size; the ".xml" version is 20k. When I render the xml into a tree displayed as a html table, the redundant namespace information for each of the document's elements causes the expansion to be 1.4MB in size! Another author [URL=http://www.infoworld.com/article/03/10/03/39FEofficerev_1.html]has also discovered[/URL] this same bloated structure.

I wonder if Microsoft's intention is to force corporations to buy expensive content management systems to deal with the bloat? It seems to me that the [URL=http://www.w3schools.com/dtd/default.asp]DTD[/URL] or [URL=http://www.w3.org/XML/Schema]XML Schema[/URL] Microsoft is using could be simplified to export to xml only the required support for the actual document and not every feature that Word can support.

You can look at the actual files I used in this discussion by downloading the ".zip" compressed file, [URL=http://www.timeoutofmind.com/code_examples/xml/word_to_xml_to_html.zip]Office 2003 Word to XML to HTML Example Code[/URL].

After uncompressing that file, you'll have a folder containing a Word document, that same Word document saved as ".xml", and an html rendering of that ".xml" file.