RE: Saxon Quirk

> This looks like a perfect opportunity for the Muenchian Method of
> finding nodes with unique properties.  
I'm glad that you said this because this is the method that I am using now
to get the entries that match [@list='yes']. I'm experimenting with other
methods because there is a need to generate a table that is grouped by other
criteria that I have not listed in my example. 

I was leaning towards generating a node-set based the variable that was
giving me the problems. This node-set would have information about how to
sort the final table. To extrapolate on my example, it might table the XML
and XSL journals under an "XML" heading, the CSS journal under an "HTML"
heading and the remaining under an "Old Language" heading. I was getting
confused trying to do this with one key or one template, so that is why I
was going the node-set route.

> Rather than go through all that malarky, and as you're using Saxon,
> you could alternatively use the saxon:distinct() extension function:
I had been originally trying to avoid processor specific extensions, but it
makes things a lot easier sometimes. Since many of the processors and the
1.1 spec support node-sets, I felt more comfortable with them. 

> P.S. Very disappointed to see 'pubs' meaning publications rather than
> the kind I could have a drink in.
Those kind of 'pubs' happen AFTER I get this project sorted out. 

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