Re: Variable and option binding proposal

/ Jeni Tennison <jeni@jenitennison.com> was heard to say:
| Norman Walsh wrote:
|> / Jeni Tennison <jeni@jenitennison.com> was heard to say:
[...]
| I was arguing that we shouldn't specify that all the variable bindings
| must come at the beginning if it doesn't matter when or if their
| values are actually computed. Let people put them where it makes sense
| to them.

Here's the case that worries me:

<p:pipeline type="main">
  <p:variable name="x" value="3"/>

  <p:xslt name="second">
    <p:with-param name="x1" select="$x"/>
    <p:input port="source">
      <p:pipe step="first" port="result"/>
    </p:input>
    <p:input port="stylesheet">
      <p:document href="http://example.com/doc.xsl"/>
    </p:input>
  </p:identity>  

  <p:variable name="x" value="4"/>

  <p:xslt name="first">
    <p:with-param name="x2" select="$x"/>
    <p:input port="source">
      <p:pipe step="main" port="source"/>
    </p:input>
    <p:input port="stylesheet">
      <p:document href="http://example.com/doc.xsl"/>
    </p:input>
  </p:identity>  
</p:pipeline>

What are the values of $x1 and $x2, respectively? How do the variables
get moved when the xslt steps are reordered.

I think it would be possible to say that it's the lexical, document
order scope that counts, even though the steps are reordered. I just
think that might give some implementors heart burn.

                                        Be seeing you,
                                          norm

-- 
Norman Walsh <ndw@nwalsh.com> | He that overvalues himself will
http://nwalsh.com/            | undervalue others, and he that
                              | undervalues others will oppress
                              | them.--Dr. Johnson