Two Types of Roster!

public class Roster {
  private Student[] students;
  private int numStudents;

  public Roster(int size) {
    students = new Student[size];
    numStudents = 0;
  }

  public void add(Student s) {
    // stub
  }

  public void remove(Student s) {
    // stub
  }

  public Student find(String email) {
    return null; //stub
  }
}

Consider the following scenarios:

1. For courses at JHU, we search more frequently than we add/remove students. Therefore, we define JhuRoster such that its find implements binary search. The add and remove methods shall keep the underlying (students) data sorted to facilitate the binary search process.

2. For MOOCs (Massive Online Open Courses) offered by "JHU Engineering for Professionals," we add/remove students much more frequently than we perform searching. Therefore, we define MoocRoster such that its find method implements linear search. Furthermore, the add and remove methods do not bother keeping the underlying (students) data sorted.

We asked students to implement JhuRoster and MoocRoster, and here are some of the design decisions they have made:

1. MoocRoster extends JhuRoster (or the other way around)
2. JhuRoster and MoocRoster are entirely different data types (or "data structures" if you like).
3. There is a Roster class where both MoocRoster and JhuRoster extend.

We will critique each design in the following section.