Research: Projects

Contact


CodeGlass

Today's software development environments and tools offer numerous ways of providing useful contextual information about the code the programmer is developing. This information is available through a number of different views, each of which has been developed to present certain information, for instance flagging buggy code, identifying the author of some code, listing to-do items still to be completed, presenting who else may be working on the code in parallel, and many others. While all this information about the code is useful, however, the number of available views is now so prevalent that they are polluting the development environment and make it difficult to work on the code. The window available for the code becomes smaller and smaller, more and more annotations are shown, and actually moving from view to view requires serious context switching.

We propose to address this problem by moving the contextual information out of the development environment and into their peripheral vision, by using Google Glass. Google Glass is designed for transient information, and the contextual information that the developers are provided about their code typically fits this nature: it is relevant for this task, file, or line of code – but only when that task, file, or line of code is ‘active’. We therefore believe that offloading the contextual onto the Google Glass offers the possibility of significantly improving a programmer’s work environment.

Naturally, our approach presents challenges of its own. While Google Glass offers extensive facilities for building “cards” through which the information can be displayed to the programmers, they cards have a limited real estate and clearly we cannot overwhelm the developer with new card after new card. The goal of this research is therefore twofold: (1) to demonstrate the feasibility of the idea, and (2) to explore ways in which to optimally use the Google Glass to present this kind of information.