How Dependencies Work


In the CMS, the site tree displays the information architecture of a site. The site tree describes hierarchical relationships between pages, components, and folders; these relationships can be described in terms of a family tree. That is, pages, components, and folders relate to each other as ancestors, parents, siblings, and children. The Reference.xml file stores this tree structure.

Pages can also be related to each other through links, navigations, components, and co-brands. These relationships are called dependencies. Files called dependency graphs maintain the dependencies between pages. Each publishing target has a corresponding dependency graph. Pages may also have dependencies that span publishing targets.

The CMS application calculates dependencies in the following circumstances:

  • A page is checked in.
  • A page is marked for publish.
  • A page is published (incremental or full).
  • A site is published (incremental or full).
  • A page is previewed.

Starting one of these processes prepares the CMS for calculating dependencies. During this preparation time, a "publish" animation is displayed and the user client is locked. The server then completes the dependency calculations in the background. Before a publish starts, the dependency calculations must be completed.

Publish times listed in the Publish Monitor and the log do not include dependency calculations. While making dependency calculations, the Publish Monitor displays "Calculating page dependencies...."