RE: [xml-dev] Binary XML == "spawn of the devil" ?

[Elliotte Rusty Harold <elharo@metalab.unc.edu>]

At 4:06 PM +0100 7/25/03, Robin LaFontaine wrote:
>There was a paper in this area by Daniel Winkowski (Mitre Corp) at 
>XML 2002 in Baltimore, with the title, "XML Sizing and Compression 
>Study for Military Wireless Data" [1]. Conclusion was as far as I 
>remember from the presentation: XML+compression is best!
>

I requested the original data for this paper when it was first 
published. The authors were unwilling to provide it. Apparently it's 
proprietary. Thus it was impossible to either verify or disprove 
their conclusions.

Evven assuming there results were accurate, it was not possible to 
determine the nature of the documents they tested. Thus I could not 
determine in what domains their results were likely to be 
significant. It was not possible to determine what characteristics of 
their data led to their results, and whether these results were 
likely to be reproducible for the types of documents I (or anyone 
else) works with.

I hope that the people reviewing submissions for the W3C workshop 
will require that all original data be made publicly and freely 
available to all interested parties so that results can be inspected, 
checked, verified, and so forth.
-- 

   Elliotte Rusty Harold
   elharo@metalab.unc.edu
   Processing XML with Java (Addison-Wesley, 2002)
   http://www.cafeconleche.org/books/xmljava
   http://www.amazon.com/exec/obidos/ISBN%3D0201771861/cafeaulaitA