RE: Better include them in the XSLT 2.0 spec (Was: Re:

> To be pedantic, my:f($x) is my:f($x) should always evaluate to true()
> in the absence of errors. But that can be taken as read.

No, given

<xsl:function name="f">
  <xsl:param name="x"/>
  <a/>
</xsl:function>

then

    f($x) is f($x)

is defined to be false. Similarly, given

<xsl:variable name="x" as="element()*">
  <xsl:for-each select="1 to 5"><a/></xsl:for-each>
</xsl:variable>

then

    count($x/.) 

is defined to be 5. (The /. causes duplicates to be eliminated: this is to
demonstrate that in this case there are no duplicates).

Certain expressions - including function calls - return distinct results
each time they are evaluated.

Michael Kay
http://www.saxonica.com/