This was an unusual project, designed as a proof-of-concept for a very large site with dozens of articles. The owner of the site specifically wanted to avoid using a database to manage content, preferring static pages, but also wanted an easy, dynamic way to generate content index pages. The solution? An XML file containing article information, processed by each article directory.

The system contains a central XML file containing the title, timestamp, and root-relative URL for each article. Each element also contains a category, which is used to generate content index pages for each directory, allowing the system to include articles in indexes directories other than the one containing the file.

Another unique feature of this system is the inclusion of a number of content flags, implemented as attributes of the main XML element for each article, allowing the articles to only be listed in directories under certain conditions. The possibilities are quite extensive when combined with the variables generated by our modular framework, including creating members-only directories, and changing the article lists based on location or URL.

Unfortunately, this application has limited appeal, but it's an excellent demonstration of using techniques and systems other than traditional CMS products to manage a website's content, while maintaining our commitment to maximise script efficiency and ease of use. And, with relatively simple modifications, this concept can be used to create an effective content management system that doesn't require a database to function.