pschuller wrote:
> I am transforming a DocBook compliant xml file into html using the DocBook 
> stylesheets distribution.  The transform runs in 2 second when I run it in 
> the oXygen  IDE.    My website is a .NET 3.5 web site.   I am doing the 
> transform programatically using an XslCompiledTransform, a MemoryStream etc.  
> That method works fine for the TOC, Introduction, and Abstract.  For the 
> sections that have <mediaobjects> in them, I immediately get a StackOverFlow 
> error ("the error").  I can display a source file that has one <mediaobject> 
> but get the error if there are two.
> 			
> Below is one of my <mediaobjects>.  There are only two in my test file.  
> (There are 350+) in the file that oXygen processes in 2 seconds.  I only have 
> to comment out the <textobject> <phrase> node to get the transform to run.

Doesn't oXygen use Saxon for XSLT transformations? There is a .NET 
version of Saxon thus you could try to use that instead of 
XslCompiledTransform. See http://saxon.sourceforge.net/
		



-- 

	Martin Honnen --- MVP XML
	http://msmvps.com/blogs/martin_honnen/