Continuous Refactoring in CI: A Preliminary Study on the Perceived Advantages and Barriers.

Our study on Continuous Refactoring presented at the IEEE International Conference on Software Maintenance and Evolution (ICSME '18) in Madrid, Spain. Preprint of the corresponding paper available at https://www.ifi.uzh.ch/seal/people/vassallo/VassalloICSME18.pdf

Transcript

1. Continuous Refactoring in CI:
   A Preliminary Study on the Perceived Advantages and Barriers.
   Carmine Vassallo, Fabio Palomba, and Harald C. Gall.
   @ccvassallo
   IEEE International Conference on Software Maintenance and Evolution, Madrid, Spain, 2018
   

   View Slide

2. Continuous Refactoring in CI:
   A Preliminary Study on the Perceived Advantages and Barriers.
   Carmine Vassallo, Fabio Palomba, and Harald C. Gall.
   @ccvassallo
   IEEE International Conference on Software Maintenance and Evolution, Madrid, Spain, 2018
   

   View Slide

3. 3
   Improved code maintainability
   Code more understandable
   Code easier to test
   Introduction of new bugs
   Lack of proper support
   Alshayeb et al., “Empirical investigation of refactoring effect on software quality”, IST 2009
   Kim et al., “A field study of refactoring challenges and benefits”, FSE 2012
   Murphy-Hill et al., “How we refactor, and how we know it.”, TSE 2011
   

   View Slide

4. Lack of Automation
   Murphy-Hill et al.,
   “How we refactor, and how we know it.”
   TSE 2011
   

   View Slide

5. CI
   Continuous
   Inspection
   Continuous
   Testing
   Continuous
   Bug Fix
   Continuous Integration
   

   View Slide

6. How Do Developers
   Perform Refactoring in CI?
   

   View Slide

7. Continuous Refactoring
   7
   

   View Slide

8. Empirical Study
   RQ1 How do developers perform refactoring in CI?
   RQ2 Do developers need to refactor continuously?
   RQ3 Which are the barriers to continuous
   refactoring?
   8
   

   View Slide

9. The Questionnaire
   9
   14 questions, 3 main topics:
   • Current Practice
   • Need for Continuous Refactoring
   • Barriers to Continuous Refactoring
   31 participants.
   

   View Slide

10. Current Practice
    

    View Slide

11. Frequency of Refactoring in CI
    11
    of our respondents look for
    refactoring opportunities at
    every change.
    %
    63
    

    View Slide

12. Motivations for Refactoring in CI
    12
    1st
    Code Quality
    Improvement
    Better Code
    Comprehension
    2nd
    Avoid to fail
    Quality Gates
    3rd
    

    View Slide

13. Motivations for Refactoring in CI
    13
    1st
    Code Quality
    Improvement
    Better Code
    Comprehension
    2nd
    3rd
    Avoid to fail
    Quality Gates
    “warnings raised by static
    analysis tools” (S11)
    “lack of test coverage” (S8)
    “code complexity is
    high” (S18)
    Adoption of Static Analysis Tools
    

    View Slide

14. The Need for
    Continuous Refactoring
    

    View Slide

15. The Need for Continuous Refactoring
    15
    of our respondents consider
    essential to continuously
    refactor.
    %
    71
    

    View Slide

16. Barriers To
    Continuous Refactoring
    

    View Slide

17. Barriers to Continuous Refactoring
    17
    Barriers
    Absence of proper test suite
    Lack of time
    Unavailability of automated tools
    

    View Slide

18. Continuous
    Refactoring
    as a
    CI (best)
    practice
    Need for
    Automated
    Refactoring
    Developer-
    oriented
    static
    analysis tools
    

    View Slide

19. 19
    Continuous Refactoring in CI:
    A Preliminary Study on the Perceived Advantages and Barriers.
    Carmine Vassallo, Fabio Palomba, and Harald C. Gall.
    @ccvassallo
    vassallo@ifi.uzh.ch
    Lack of Automation
    Murphy-Hill et al.,
    “How we refactor, and how we know it.”
    TSE 2011
    Frequency of Refactoring in CI
    X
    of our respondents look for
    refactoring opportunities at
    every change.
    %
    63
    Barriers to Continuous Refactoring
    X
    Barriers
    Absence of proper test suite
    Lack of time
    Unavailability of automated tools
    

    View Slide