| CompSci 308 Spring 2026 |
Advanced Software Design and Implementation |
The only way to make the deadline — the only way to go fast — is to keep the code as clean as possible at all times. — Robert Martin on LeBlanc's Law: "Later equals Never."
Many view software design as a craft, learned best through practice that is guided by experienced mentors, where the best lessons are learned by making mistakes rather than there being one right way to do it. In this respect it is more like art than science. The arts, as well as architecture and engineering, benefit from subjecting work to constant review by both peers and mentors as a means of providing fast feedback which allows students to learn quickly from their mistakes.
This course focuses on the design of software by implementing successively larger projects in successively larger teams. As such, students should focus on improving their program's flexibility while communicating with other students to fit their code into a larger whole. Projects given are intended to be under-specified, so that their design and functionality must be justified and improved through iterative feedback. My goal is for you to be able to turn vague and ambitious specifications into a system design that can be built, maintained, and extended, prioritizing the features to be built first, then refining and extending the project through multiple draft releases.
This course is geared towards students who enjoy software development and abstract thinking and want to learn practical ways to better design and code their projects. If you find your efforts are not being well directed towards learning, please let the Teaching Team know so we can help remedy the situation — the goal is not to work you simply for work's sake!
Required Background. It is assumed that you have previously programmed extensively in Java (e.g., at least the level of CompSci 201 at Duke). If you do not have such experience programming, this probably is not the right course for you. See the Teaching Team if you have questions.
Help us improve the course. We welcome your feedback about all aspects of the course as the Computer Science department at Duke University aims to excel in education and research.
Help us improve the website. Please inform us of any broken links or otherwise misleading or confusing content and we will get it fixed as soon as possible.
Our goal is to adapt the subject and delivery of this course to meet your needs while maintaining our high standards.