Code Review is an Architectural Necessity - SATURN 2016

There's a wealth of material on code review from a code quality standpoint, tracking a host of metrics and generating enough Big Data to employ a small army of analysts at some companies, I'm sure. But introducing code review at the architecture stage seems to be rarely done, maybe even sufficiently rare to qualify as novel. In this presentation, I intend to focus on some quality attributes valued by a team that conscientiously conducts code reviews, and that code review enables, but not guarantees, three main attributes in the team's systems' architecture: maintainability, compliance, and security.

I posit based on my own meandering experience across several projects, open and proprietary, that these quality attributes are enabled through code review and saved from being a poorly timed afterthought or patch onto an architecture.

Transcript

1. Code review is an architectural necessity Colin Dean @colindean 1
2. @ColinDean Software Engineer Organizer, Abstractions.io Wearer of many hats 2
3. My words are my own and not my employer(s), past or present. Please save questions until the end of the presentation. 3
4. Agenda • Quick anecdote • What is code review? • What problems does code review solve? • Quality attributes code review ensures • Tips for code reviews • Limitations 4
6. Agenda • Quick anecdote • What is code review? • What problems do code review solve? • Quality attributes code review ensures • Tips for code reviews • Limitations 6
7. What is code review? 7
8. Code review is the process by which those who maintain a software codebase evaluate a proposed change to that codebase, regardless of the source of the proposed change. 8
9. Code review is systematic examination of computer source code. 9 Code Review, Wikipedia
10. Peer Review 10
11. Code Review 11
12. Code Review Vocabulary • Change - an individual unit of work altering what exists • Submission - a collection of changes • Submitter - the person proposing the submission • Reviewer - the people evaluating the submission • Annotation - remarks or ratings bestowed upon the submission 12
13. The submitter proposes changes in a submission, which is evaluated by a reviewer, who annotates or accepts it. 13
14. Inspection Team review Walkthrough Pair programming Peer deskcheck, passaround Ad-hoc review Wiegers’ peer review formality spectrum 14 Least formal Most formal
15. Most formal Least formal Inspection Team review Walkthrough Pair programming Peer deskcheck, passaround Ad-hoc review Wiegers’ peer review formality spectrum 15
17. Agenda • Quick anecdote • What is code review? • What problems does code review solve? • Quality attributes code review ensures • Tips for code reviews • Limitations 17
18. Code review solves two major problems. 18 Aside from the primary goal of reducing defects,
19. Mental model synchronization Code review solves 19
22. Close enough Need guidance On target 22
23. Tribal knowledge development Code review solves 23
24. Michael Keeling
 Creating an Architecture Oral History, SATURN 2012 “Architecture oral history requires that the team is both willing and able to retell the stories and keep the oral history alive.” 24
25. Write it down. Make it searchable. Code review forces us to 25
26. Agenda • Quick anecdote • What is code review? • What problems does code review solve? • Quality attributes code review ensures • Tips for code reviews • Limitations 26
27. Maintainability Code review ensures 27
28. Maintainability • Learnability • Understandability • Serviceability Code review drives 28
29. Learnability • Developing Code • Patterns & Conventions • Risks & Goals • Developing People • Common Vocabulary • Teaching Moments Maintainability Learnability Understandability Serviceability Code review drives 29
30. Learner Expert Coding Reviewing Coding Reviewing Synchronous Pairing & Teaching Exemplary Reading Constructively Critical Evaluation Serendipitous Evaluation of Example Maintainability Learnability Understandability Serviceability 30
31. Understandability • Establishes common yet evolving mental model • Builds confidence in direction and design decisions • Builds tribal knowledge • Bonus: Enables elevator pitch Maintainability Learnability Understandability Serviceability Code review drives 31
32. Serviceability • Exposes addressable “gotchas” • Exposes end-user interaction points • Establishes consensus on supported workflows Maintainability Learnability Understandability Serviceability Code review drives 32
33. Serviceability • Exposes addressable “gotchas” • Exposes end-user interaction points • Establishes consensus on supported workflows Maintainability Learnability Understandability Code review drives 32
34. Linus’s Law “Given enough eyes, all bugs are shallow.” 33
35. Maintainability ✓Learnability ✓Understandability ✓Serviceability Code review drives 34
36. First programming job out of school - B2B imprinting company if($customer == “spacely_sprockets”) { do_something(); }
 else { cry(); } • Version control! • No code review tooling or process • Minimal pairing • Continous integration easily circumvented 35
37. Lack of code review 36 Lost Opportunities
38. Lack of code review 37 Lost Opportunities Lost Revenue
39. Lack of code review Lost Opportunities Lost Revenue Lost Job 38
40. Compliance Code review ensures 39
41. Compliance • Accessibility • Auditability • Idiomaticity Code review drives 40
42. Second job out of school - Consulting • Lone wolf working alongside other lone wolves • No version control in proprietary software with custom “IDE” a.k.a. textarea. • Last modified and modifier only • No process of our own 41
43. First professional code review experience was group review • Subcontractor on government project, 2010-2012 • Lone SME on platform • Borland StarTeam + in house review system • My tools for version control integration • Weekly merge window • Round robin inspection 42
45. Not a pleasant experience • Three to four hour weekly round robin inspection • Cutthroat mixture of competing contractors, subcontractors, and employees • Embarrassment galore ‛ Not a learning environment • Immediate defensive posture • “Merge next week” = you failed, possibly delayed project 44
46. $1,450 per hour 45
47. $1,450 per hour $5,800 per weekly meeting 46
48. $1,450 per hour $5,800 per weekly meeting $290,000 per year 47
49. Effects? • Waste • “Get this over with.” • Obstructionism • Plenty of bugs • “I’ll fix that mistake later.” 48
50. Missed opportunities • Accessibility expert was most vocal • Project manager was vocal on contractual and HF matters ➡ Both could have reviewed asynchronously • Project was behind ➡ Too many people could say No 49
51. Security Code review ensures 50
52. Security • Spot vulnerabilities • Teach best practices • Filter unnecessary code • YAGNI Code review drives 51
53. Reviewers are like your lawyer Screening and recommending actions to minimize risk, avoid preventable mistakes 52
54. Agenda • Quick anecdote • What is code review? • What problems does code review solve? • Quality attributes code review ensures • Tips for code reviews • Limitations 53
55. When should you integrate code review? 54
56. Context • Project • Technical 55
57. Keep reviews informal and short. 56
58. Tips for thorough code review 57
59. Tips for thorough code review • Devote time 57
60. Tips for thorough code review • Devote time • Accept debt 57
61. Tips for thorough code review • Devote time • Accept debt • Identify churn 57
62. Tips for thorough code review • Devote time • Accept debt • Identify churn • Minimize pedantry 57
63. Tips for thorough code review • Devote time • Accept debt • Identify churn • Minimize pedantry • Make progress 57
64. Major things we look for • Algorithmic complexity • Exception & error handling • Exception, class, & variable naming • Logging sufficiency & level • Style conformation (automate!) • Long lines & methods • Readability • Single purpose per commit 58
65. Most importantly 59
66. Most importantly Does it work? 59
67. Most importantly Does it work? Is it tested? 59
68. Agenda • Quick anecdote • What is code review? • What problems does code review solve? • Quality attributes code review ensures • Tips for code reviews • Limitations 60
69. Analyze dynamic structures Code review cannot 61
70. Go on endlessly Code review cannot 62
71. Solve political problems Code review cannot 63
72. Agenda • Quick anecdote • What is code review? • What problems do code review solve? • Quality attributes code review ensures • Tips for code reviews • Limitations 64
73. Code Review 65
74. Code Review is systemic examination of proposed changes to a codebase. 65
75. Code Review is systemic examination of proposed changes to a codebase. solves mental model synchronization and tribal knowledge development. 65
76. Code Review is systemic examination of proposed changes to a codebase. solves mental model synchronization and tribal knowledge development. ensures maintainability, compliance, & security. 65
77. Code Review is systemic examination of proposed changes to a codebase. solves mental model synchronization and tribal knowledge development. ensures maintainability, compliance, & security. must be short, thorough, and automated where possible. 65
78. Code Review is systemic examination of proposed changes to a codebase. solves mental model synchronization and tribal knowledge development. ensures maintainability, compliance, & security. must be short, thorough, and automated where possible. will not solve all human problems, but some is better than none. 65
80. 1,500+ software professionals in Pittsburgh in August abstractions.io @abstractionscon 66
81. @ColinDean github.com/
 colindean/talks speakerdeck.com/ colindean 67
82. FIN 68
83. Attributions • Westminster College picture: https://www.flickr.com/photos/westminstercollege/15759678054/in/album-72157649340620016/ • RMU picture: http://cfbarchitects.com/higher-education/selected-projects/academic-buildings-libraries-learning-commons/robert- morris-university/ • Pittsburgh picture: probably Dave DiCello • On switch https://openclipart.org/detail/180085/switch-on • Off switch https://openclipart.org/detail/180084/switch-off • “Their first code review” http://classicprogrammerpaintings.tumblr.com/post/142702963264/their-first-code-review-william- frederick • Bass, Len; Paul Clements, and Rick Kazman. Software Architecture in Practice. Addison Wesley, 2013. • Wiegers, Karl E. Peer Reviews in Software. Addison Wesley, 2012. • Cohen, Jason, Steven Teleki, and Eric Brown. Best Kept Secrets of Peer Code Review. Smart Bear Software, 2006. • Wilhelm, Alex and Alexia Tsotsis. Julie Ann Horvath Describes Sexism and Intimidation behind Her Github Exit. TechCruch, 2014 March 15. Retrieved 2016 April 26. http://techcrunch.com/2014/03/15/julie-ann-horvath-describes-sexism-and-intimidation- behind-her-github-exit/ • and others mentioned in the slides 69
84. No, really. Fin. Srsly. 70
85. Third out of school and current job - Engineering • Highly disciplined team using Java, Scala, and Groovy • Git + Gerrit • Constructively critical feedback • No criticism without alternative solution and reasoning • Wide experience range: 1-2 yrs to 25+ yrs • Team split in late 2014, I was asked to be tech lead 71
86. Github Enterprise in 2016 • All new projects • Same workflow as public Github 72
87. Code Review Tools Used Haven’t Used Like Dislike ̣Github ̣Gerrit ̣Gitlab •Gitbucket •BitBucket ̣StarTeam •Phabricator •git-assess 73