Ah!  Thank you for the correction.  One feature of this mailing list I
appreciate is how sloppy thinking is put to the sword!

>  From: David Carlisle
> 
>  > I believe this works as follows:
>  more or less, yes.
>  
>  
>  > 1. the select in copy-of returns a node-set in document 
>  order.  At this
>  > point the node set has duplicate attribute names.
>  
>  As I've said so often on this list that even Jeni's started saying it,
>  set's don't have an order. So the select just selects the set 
>  of nodes.
>  
>  Howver copy-of, in common with all other result building operations
>  in xslt, processes the nodes in the set in document order, and so:
>  
>  > 2. when copy-of applies these attributes to the output element, the
>  > duplication is removed by succeeding attribute nodes 
>  overwriting the values
>  > of any attributes that are already present.
>  
>  
>  David
>  

 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list