Until now, we've considered each activity to always correspond to a single screen in our application. We used fragments only to represent subsections within each screen. As an example, let's think back to the way we constructed our book browsing application. In the case of a wide-display device, our application uses a single activity containing two fragments. One fragment displays the list of book titles, and the other fragment displays the description of the currently selected book. As both of these fragments appear onscreen at the same time, we display and manage them from a single activity. However, in the case of a portrait-oriented handset, we chose to display the book list and the book description on separate screens. The reason for this is that the two fragments do not appear onscreen at the same time; we manage them in separate activities.
Surprisingly, in both cases our application performed the same task. The only difference is how much information...