8.3 Scripted Paths

Polle Zellweger generalizes the notion of guided tours in her description of scripted paths [Zel89]. She sees her path mechanism as a fundamental solution to the two classical hypertext problems of user disorientation and the additional cognitive overhead needed to create and choose among links. Obviously, paths should not replace browsing as main hypertext navigation mechanism, but rather augment it. Zellweger distinguishes between:

• Sequential paths, i.e., ordered sequences of entries.
• Branching paths, where the reader has to manually choose the next node among a selection of different links.
• Conditional paths, where the branching is controlled by the system, e.g., by selecting the next node based on the previous actions of the reader.

Zellweger lists different representations for paths. She distinguishes between "embedded link" representation, "separate textual" representation, and "separate graphical" representation (figure I.31).

Figure I.31 Alternative path representations

The top of figure I.31 shows a sequential path in the three different representations. The separate textual representation (in the middle) corresponds to a description of the path in a programming language. The bottom of figure I.31 shows a conditional path in the three different representations. All six path representations of figure I.31 are based on the same underlying hypertext structure consisting of the four linked nodes A, B, C and D (figure I.32).

.
Figure I.32 Basic hypertext structure of figure I.31

A path as defined by Zellweger has to offer some sort of playback control. This means that the user needs an easy way to follow a path. This can either be a single stepping control, allowing the user to step from one node of the path to the other, or automatic playback control. Users may even get full browsing control allowing them to visit the nodes of a path in arbitrary order.

Zellweger implemented her ideas in a hypermedia system called Scripted Documents[Zel89]. Scripted Documents provides paths as its sole linking mechanism. The idea is to link documents of different types, as e.g., text, voice, music, video, by paths. A script consists of three parts:

• A set of documents, which are normally preexisting files.
• A set of script entries, associating locations within the documents with actions, as to play back a voice fragment or an animation.
• A path specification, describing the route that the author intends the reader to take.

Scripted paths in Scripted Documents offer the types of playback control (single stepping, automatic playback, browsing) as described above. They also allow the reader to preview a path and to jump directly to an entry.

Zellweger describes a flexible system that vastly extends the ideas of guided tours as described in the two previous sections. By allowing paths to be specified in a programming language, it becomes straightforward to implement conditional paths and integrate any type of document into the path. Zellweger gives us a powerful tool to build hypermedia documents of all sorts, but with the flexibility of scripted paths we also get the capability to build the most complicated, tangled webs where the reader gets hopelessly lost. It is therefore up to the author to use the scripted path mechanism with caution, and to build hyperdocuments that prevent the user from becoming disoriented and frustrated by too much cognitive overhead needed to create and choose among links.

The final system presented in this chapter, Footsteps, is again much simpler than Scripted Paths, but is implemented on a much wider available platform, the web.