Transparency, holophrasting, and automatic layout applied to control structures for visual dataflow programming languages

摘要

Control flow within dataflow visual programming languages such as Prograph is difficult to understand due to hidden dependencies created by the control structures that desperate code blocks into independent views. We have created a model arising from research on ReactoGraph -- a derivative of Prograph -- that allows code in a general class of dataflow programming languages to be controlled based on nested graph structures. These structures, aimed at removing hidden dependencies, provide a reasonable alternative to Prograph's case control structures. However, they have also been found to introduce some usability problems such as low tolerance to code layout modification and difficulties scaling the number of code elements as nesting depth increases. In this paper we present a novel use of various software visualization techniques applied to these nested control structures in order to enable effective modification and scalability of code for improved comprehension and programming performance.