The Drupal codebase is read and discussed by developers and maintained as different versions using a concurrent version system (CVS). There, core and contributed modules, patches of new code, branches of code representing different versions of core and contributed modules, function not as a strictly linear progression of Drupal numbered versions, but rather as a multilinear hypertext of paths: some successful implementations which made it into the core or contributed modules permanently, part of the continued evolution of the product; others, false starts or modules in languished development. All-in-all, a mega hypertextual narrative more complex than any hyperfiction text; intertextuality made more transparent.

Accompanying the primary text of Drupal itself are the discussions of the developers, a decision-making discourse which annotates the incorporation of new code into Drupal. As with most open source development communities, code goes through a rigorous peer review process where developers discuss implementation, design, and code efficiency (Raymond, Taylor and Riley). One might assume that the developers--along with ideas from regular users--make all the decisions about the code. However, another participant is recognizable in the discussions: Drupal. Drupal is often personified, with developers regularly pointing to what Drupal needs, what is right for Drupal, or what works for Drupal. Drupal, then, is represented in the vision for its construction that is both unwritten and represented by the text of the code itself and the many discussions about code and application development. This unique situation exists mainly in open source CMS communities. The software has needs to support its own development, fostering the relationship of the code as user of its own application as a voice within the community it also serves.