Professional Background

  1. Are you a professional or part-time software developer?
    • A. Professional
    • B. Part-Time
  2. How large is your company?
    • A. Up to 10 employees
    • B. 11-50 employees
    • C. 51-100 employees
    • D. > 100 employees
  3. How many projects have you worked on?
    • A. Up to 5 projects
    • B. 6-10 projects
    • C. 11-15 projects
    • D. > 15 projects
  4. How many years of Java programming experience do you have?
    • A. < 2 years
    • B. 2-5 years
    • C. 6-10 years
    • D. > 10 years
  5. How many years of continuous integration (CI, e.g., Travis) experience do you have?
    • A. < 2 years
    • B. 2-5 years
    • C. 6-10 years
    • D. > 10 years

Approach Motivation

  1. If you ever used CI, how often does your team trigger the CI build of your projects?
    • A. minutes
    • B. hours
    • C. days
    • D. weeks
    • Any comments? ______
  2. If you ever used CI, do you think the CI build is time-consuming?
    • A. Absolutely Yes
    • B. Probably Yes
    • C. Not Sure
    • D. Probably Not
    • E. Absolutely Not
    • Any comments? ______
  3. If there is an automated tool that can predict CI build outcomes (i.e., either passed or failed) without running builds (i.e., a project is built only when the predicted build outcome is failed), do you think it would be useful for CI-based software development?
    • A. Absolutely Yes
    • B. Probably Yes
    • C. Not Sure
    • D. Probably Not
    • E. Absolutely Not
    • Any comments? ______
  4. Why do you think CI build outcome prediction would be useful?
    • A. Obtain quick feedback of CI builds
    • B. Save time overhead of CI builds
    • C. Reduce resource consumption in CI servers
    • D. Accelerate software development
    • E. Others ______
  5. Why do you think CI build outcome prediction would not be useful?
    • A. Lack of prediction accuracy
    • B. Lack of explainability
    • C. Incorrect prediction may delay the discovery of bugs (e.g., compilation error or test failure)
    • D. Incorrect prediction may increase the difficulty of bug fixing
    • E. Others ______