<?xml version="1.0" encoding="utf-8"?>
<rss xmlns:content="http://purl.org/rss/1.0/modules/content/" version="2.0">
  <channel>
    <title><![CDATA[Content with Style - Comments on Problem: Dependencies]]></title>
    <link>http://www.contentwithstyle.co.uk/feeds/rss/comments/115</link>
    <description><![CDATA[]]></description>
    <pubDate>Tue, 06 Jan 2009 07:18:43 +0000</pubDate>
    <generator>Zend_Feed</generator>
    <docs>http://blogs.law.harvard.edu/tech/rss</docs>
    <item>
      <title><![CDATA[Content with Style - Comment #1 on Problem: Dependencies]]></title>
      <link>http://www.contentwithstyle.co.uk/content/problem-dependencies/#comment-1168</link>
      <guid>http://www.contentwithstyle.co.uk/content/problem-dependencies/#comment-1168</guid>
      <description><![CDATA[Pascal,<br />
<br />
I have long used an XML based topic map or page tree. I also used a depth attribute. But I also found that other &#8220;meta data&#8221; such as adding an id, created/updated dates to the pages were useful for me. That also allows me to have pages that introduce other pages to automatically take on the last updated date of the most recently updated child page.<br />
<br />
I don&#8217;t know if that is what you are looking for?<br />
<br />
In my case the dependency is strictly hierarchical. I don&#8217;t know if that is the case for you.<br />
<br />
One though if your pages depend on pages throughout the tree, is to give each page an id, and perhaps if easy to manage have a list of dependent ids per page. Then, your pagetree has all this information at hand. With XSLT you can easily reference other pages if needed (you might even be able to use keys to speed up the lookup&#8230;)]]></description>
      <content:encoded><![CDATA[Pascal,<br />
<br />
I have long used an XML based topic map or page tree. I also used a depth attribute. But I also found that other &#8220;meta data&#8221; such as adding an id, created/updated dates to the pages were useful for me. That also allows me to have pages that introduce other pages to automatically take on the last updated date of the most recently updated child page.<br />
<br />
I don&#8217;t know if that is what you are looking for?<br />
<br />
In my case the dependency is strictly hierarchical. I don&#8217;t know if that is the case for you.<br />
<br />
One though if your pages depend on pages throughout the tree, is to give each page an id, and perhaps if easy to manage have a list of dependent ids per page. Then, your pagetree has all this information at hand. With XSLT you can easily reference other pages if needed (you might even be able to use keys to speed up the lookup&#8230;)]]></content:encoded>
      <pubDate>Fri, 28 Jul 2006 07:57:32 +0000</pubDate>
    </item>
    <item>
      <title><![CDATA[Content with Style - Comment #2 on Problem: Dependencies]]></title>
      <link>http://www.contentwithstyle.co.uk/content/problem-dependencies/#comment-1169</link>
      <guid>http://www.contentwithstyle.co.uk/content/problem-dependencies/#comment-1169</guid>
      <description><![CDATA[Anup, I&#8217;m afraid it&#8217;s not as simple as that. Navigations could refer to ANY set of nodes expressed through ANY XPATH query, and not stricly hierachical like in your query.<br />
<br />
Therefore it&#8217;s impossible to determine with other pages will be dependencies by just keeping track of  ID&#8217;s that are lower in the hierarchy of the tree.<br />
<br />
Also, given that the rendering of the HTML output is happening through XSLT, how can the XSLT itself keep track of what&#8217;s the latest element inserted without being called actually? The whole purpose of this is to figure out the dependencies in order to avoid unnecessary rendering, you see?]]></description>
      <content:encoded><![CDATA[Anup, I&#8217;m afraid it&#8217;s not as simple as that. Navigations could refer to ANY set of nodes expressed through ANY XPATH query, and not stricly hierachical like in your query.<br />
<br />
Therefore it&#8217;s impossible to determine with other pages will be dependencies by just keeping track of  ID&#8217;s that are lower in the hierarchy of the tree.<br />
<br />
Also, given that the rendering of the HTML output is happening through XSLT, how can the XSLT itself keep track of what&#8217;s the latest element inserted without being called actually? The whole purpose of this is to figure out the dependencies in order to avoid unnecessary rendering, you see?]]></content:encoded>
      <pubDate>Fri, 28 Jul 2006 08:18:46 +0000</pubDate>
    </item>
  </channel>
</rss>
