Authorware is based on a flowchart metaphor rather than a card/stack or book/page metaphor, and its architecture does not naturally lend itself to applications based on page-turning. Although forward and backward paging structures can be created (both textbooks and magazines have been implemented using Authorware), they are not native to Authorware's paradigm.
Authorware is not a text search and retrieval engine, although it can be linked to one if necessary. There are no search or sort routines in the current version (1.0 for Windows, 1.5.3 for Macintosh).
Authorware does not contain as wide a range of data structures as traditional programming languages. For example, while there is support for single-dimensional arrays, there are no multi-dimensional arrays.
Authorware applications are interpreted rather than compiled. While Authorware's iconic structure runs faster than most text-based interpreted languages, the packaged files created are not object code, strictly speaking.