Refactoring the Technical Interview

Is your technical interview optimized? Could it use a refactor?

Do you ask candidates questions that could actually influence your production codebase? Or, do you ask them about problems that are solved and standardized? Or worse, contrived and trivial?

Leverage your production code and commit history to make your next interview more effective in identifying the right hire. All while having more fun.

Transcript

1. THE TECHNICAL INTERVIEW REFACTORING
2. WHY EVEN BOTHER? REFACTORING THAT IS
3. MARK SIEMERS RUBYCONF 2018 THE HIGH COSTS OF BAD HIRING DECISIONS ▸ Time ▸ Money ▸ Resources ▸ Opportunity Cost ▸ Reputation ▸ Morale ▸ Technical Debt REFACTORING THE TECHNICAL INTERVIEW
4. MARK SIEMERS RUBYCONF 2018 THE HIGH COSTS OF BAD HIRING DECISIONS ▸ Time ▸ Money ▸ Resources ▸ Opportunity Cost ▸ Reputation ▸ Morale ▸ Technical Debt HOW MUCH IS IT COSTING YOU? REFACTORING THE TECHNICAL INTERVIEW
5. IS ANYTHING WRONG?
6. DO NEW CANDIDATES WRITE CODE DURING THEIR INTERVIEW? Joel Spolsky - The Joel Test (#11)
7. MARK SIEMERS RUBYCONF 2018 ANY CODING QUESTION IS FINE, RIGHT? ▸ Find min (or max) in a list ▸ Implement reversing a string ▸ Implement a sorting algorithm ▸ FizzBuzz (not actually FizzBuzz, but any contrived algorithm) ▸ “Now can you do it recursively?” REFACTORING THE TECHNICAL INTERVIEW
8. IN A CARPENTRY INTERVIEW… ASK ANY QUESTION
9. MARK SIEMERS RUBYCONF 2018 EQUIVALENT CARPENTRY INTERVIEW QUESTIONS REFACTORING THE TECHNICAL INTERVIEW
10. MARK SIEMERS RUBYCONF 2018 EQUIVALENT CARPENTRY INTERVIEW QUESTIONS ▸ How do you build a drill? REFACTORING THE TECHNICAL INTERVIEW
11. MARK SIEMERS RUBYCONF 2018 EQUIVALENT CARPENTRY INTERVIEW QUESTIONS ▸ How do you build a drill? ▸ How do you build a circular saw? REFACTORING THE TECHNICAL INTERVIEW
12. MARK SIEMERS RUBYCONF 2018 EQUIVALENT CARPENTRY INTERVIEW QUESTIONS ▸ How do you build a drill? ▸ How do you build a circular saw? ▸ Imagine a structure where ▸ Every 3rd plank is birch ▸ Every 5th plank is maple ▸ If the plank is divisible by 3 and 5, it must be maple and birch ▸ how many maple/birch planks in a 3 story structure? REFACTORING THE TECHNICAL INTERVIEW
13. MARK SIEMERS RUBYCONF 2018 EQUIVALENT CARPENTRY INTERVIEW QUESTIONS ▸ How do you build a drill? ▸ How do you build a circular saw? ▸ Imagine a structure where ▸ Every 3rd plank is birch ▸ Every 5th plank is maple ▸ If the plank is divisible by 3 and 5, it must be maple and birch ▸ how many maple/birch planks in a 3 story structure? REFACTORING THE TECHNICAL INTERVIEW WTF???
14. WHY???
15. WHY??? DO YOU WANT ME TO USE THIS ON THE JOB?
16. MARK SIEMERS RUBYCONF 2018 ISSUES WITH THESE QUESTIONS (CARPENTRY) ▸ Solved ▸ Building drills and circular saws was solved long ago ▸ Standardized ▸ These tools are trustworthy and widely available ▸ Trivial ▸ Last example is an arbitrary and trivial task REFACTORING THE TECHNICAL INTERVIEW
17. MARK SIEMERS RUBYCONF 2018 ISSUES WITH THESE QUESTIONS (SOFTWARE DEVELOPMENT) REFACTORING THE TECHNICAL INTERVIEW
18. MARK SIEMERS RUBYCONF 2018 ISSUES WITH THESE QUESTIONS (SOFTWARE DEVELOPMENT) ▸ Solved REFACTORING THE TECHNICAL INTERVIEW
19. MARK SIEMERS RUBYCONF 2018 ISSUES WITH THESE QUESTIONS (SOFTWARE DEVELOPMENT) ▸ Solved ▸ Quicksort ~ 1959 (Pre-dates cordless electric drill by about 10 years) REFACTORING THE TECHNICAL INTERVIEW
20. MARK SIEMERS RUBYCONF 2018 ISSUES WITH THESE QUESTIONS (SOFTWARE DEVELOPMENT) ▸ Solved ▸ Quicksort ~ 1959 (Pre-dates cordless electric drill by about 10 years) ▸ Merge Sort ~ 1945 (The year World War II ended) REFACTORING THE TECHNICAL INTERVIEW
21. MARK SIEMERS RUBYCONF 2018 ISSUES WITH THESE QUESTIONS (SOFTWARE DEVELOPMENT) ▸ Solved ▸ Quicksort ~ 1959 (Pre-dates cordless electric drill by about 10 years) ▸ Merge Sort ~ 1945 (The year World War II ended) ▸ Radix Sort ~ 1887 (About 35 years before the electric circular saw) REFACTORING THE TECHNICAL INTERVIEW
22. MARK SIEMERS RUBYCONF 2018 ISSUES WITH THESE QUESTIONS (SOFTWARE DEVELOPMENT) ▸ Solved ▸ Quicksort ~ 1959 (Pre-dates cordless electric drill by about 10 years) ▸ Merge Sort ~ 1945 (The year World War II ended) ▸ Radix Sort ~ 1887 (About 35 years before the electric circular saw) ▸ Standardized REFACTORING THE TECHNICAL INTERVIEW
23. MARK SIEMERS RUBYCONF 2018 ISSUES WITH THESE QUESTIONS (SOFTWARE DEVELOPMENT) ▸ Solved ▸ Quicksort ~ 1959 (Pre-dates cordless electric drill by about 10 years) ▸ Merge Sort ~ 1945 (The year World War II ended) ▸ Radix Sort ~ 1887 (About 35 years before the electric circular saw) ▸ Standardized ▸ What language or library doesn’t have sort, max, min, reverse, etc. built in? REFACTORING THE TECHNICAL INTERVIEW
24. MARK SIEMERS RUBYCONF 2018 ISSUES WITH THESE QUESTIONS (SOFTWARE DEVELOPMENT) ▸ Solved ▸ Quicksort ~ 1959 (Pre-dates cordless electric drill by about 10 years) ▸ Merge Sort ~ 1945 (The year World War II ended) ▸ Radix Sort ~ 1887 (About 35 years before the electric circular saw) ▸ Standardized ▸ What language or library doesn’t have sort, max, min, reverse, etc. built in? ▸ Studied REFACTORING THE TECHNICAL INTERVIEW
25. MARK SIEMERS RUBYCONF 2018 ISSUES WITH THESE QUESTIONS (SOFTWARE DEVELOPMENT) ▸ Solved ▸ Quicksort ~ 1959 (Pre-dates cordless electric drill by about 10 years) ▸ Merge Sort ~ 1945 (The year World War II ended) ▸ Radix Sort ~ 1887 (About 35 years before the electric circular saw) ▸ Standardized ▸ What language or library doesn’t have sort, max, min, reverse, etc. built in? ▸ Studied ▸ Hundreds of publications dedicated to the coding interview REFACTORING THE TECHNICAL INTERVIEW
26. MARK SIEMERS RUBYCONF 2018 ISSUES WITH THESE QUESTIONS (SOFTWARE DEVELOPMENT) ▸ Solved ▸ Quicksort ~ 1959 (Pre-dates cordless electric drill by about 10 years) ▸ Merge Sort ~ 1945 (The year World War II ended) ▸ Radix Sort ~ 1887 (About 35 years before the electric circular saw) ▸ Standardized ▸ What language or library doesn’t have sort, max, min, reverse, etc. built in? ▸ Studied ▸ Hundreds of publications dedicated to the coding interview ▸ Trivial (FizzBuzz, Bubble Sort, etc.) REFACTORING THE TECHNICAL INTERVIEW
27. WHY DO WE ASK THESE QUESTIONS?
28. I’M NOT SURE…
29. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW OPTIMAL FOR ▸ Algorithm Writers ▸ Procedure Implementers
30. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW OPTIMAL FOR ▸ Object Architects ▸ System Designers ▸ Application Developers ▸ Many, many other roles NOT OPTIMAL FOR ▸ Algorithm Writers ▸ Procedure Implementers
31. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW While the skills on the left are valuable, Often, the skills on the right are more valuable. OPTIMAL FOR ▸ Object Architects ▸ System Designers ▸ Application Developers ▸ Many, many other roles NOT OPTIMAL FOR ▸ Algorithm Writers ▸ Procedure Implementers
32. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW TWO QUESTION LITMUS TEST FOR YOUR INTERVIEW
33. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW TWO QUESTION LITMUS TEST FOR YOUR INTERVIEW ▸ Would you be happy or upset/confused if someone did this on the job? ▸ What if someone actually implemented bubble sort on the job? FizzBuzz?
34. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW TWO QUESTION LITMUS TEST FOR YOUR INTERVIEW ▸ Would you be happy or upset/confused if someone did this on the job? ▸ What if someone actually implemented bubble sort on the job? FizzBuzz? ▸ Which book would best prepare someone for your interview …
36. THIS ONE?
37. THIS ONE? OR THIS ONE?
38. ANSWERED YES TO EITHER QUESTION?
39. ANSWERED YES TO EITHER QUESTION? CONSIDER A SLIGHTLY DIFFERENT APPROACH.
40. WHAT IF I STILL NEED AN ALGORITHM WRITER?
41. WHAT IF I STILL NEED AN ALGORITHM WRITER? OR I’M NOT SURE OF THE RIGHT BALANCE?
42. WHAT IS THE POINT OF AN INTERVIEW?
43. WHAT IS THE POINT OF AN INTERVIEW? TO SEE IF SOMEONE CAN DO THE JOB
44. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW JOB EXPECTATIONS
45. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW ▸ Read Code JOB EXPECTATIONS
46. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW ▸ Read Code ▸ Understand It JOB EXPECTATIONS
47. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW ▸ Read Code ▸ Understand It ▸ Make it better JOB EXPECTATIONS
48. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW ▸ Read Code ▸ Understand It ▸ Make it better JOB EXPECTATIONS JOB INTERVIEW EXPECTATIONS
49. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW ▸ Read Code ▸ Understand It ▸ Make it better ▸ Read Code JOB EXPECTATIONS JOB INTERVIEW EXPECTATIONS
50. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW ▸ Read Code ▸ Understand It ▸ Make it better ▸ Read Code ▸ Explain It JOB EXPECTATIONS JOB INTERVIEW EXPECTATIONS
51. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW ▸ Read Code ▸ Understand It ▸ Make it better ▸ Read Code ▸ Explain It ▸ Make it better JOB EXPECTATIONS JOB INTERVIEW EXPECTATIONS
52. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW ▸ Read Code ▸ Understand It ▸ Make it better ▸ Read Code ▸ Explain It ▸ Make it better JOB EXPECTATIONS JOB INTERVIEW EXPECTATIONS What about writing new code?
53. “IT’S HARDER TO READ CODE THAN TO WRITE IT.” Joel Spolsky, A Fundamental Law of Programming
54. “CODE IS READ MANY MORE TIMES THAN IT IS WRITTEN.” Sandi Metz
55. "THE RATIO OF TIME SPENT READING VERSUS WRITING IS WELL OVER 10 TO 1.” Robert C. Martin, Clean Code
56. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW READING CODE VS WRITING CODE
57. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW READING CODE VS WRITING CODE ▸ Do you want to ask about the part of the job that is: ▸ 90% of the technical work and more difficult ▸ 10% of the technical work and easier
58. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW READING CODE VS WRITING CODE ▸ Do you want to ask about the part of the job that is: ▸ 90% of the technical work and more difficult ▸ 10% of the technical work and easier ▸ Even with “new” code on the job, do you expect: ▸ Starting from scratch without referencing any other code? ▸ Referencing examples and applying them to the problem at hand?
59. GREAT, BUT WHAT CODE SHOULD THEY READ?
60. GREAT, BUT WHAT CODE SHOULD THEY READ? YOURS
61. LET’S TRY IT.
62. READ CODE. EXPLAIN IT. MAKE IT BETTER.
63. READ CODE. EXPLAIN IT. MAKE IT BETTER. WHAT IS THIS CODE DOING?
64. READ CODE. EXPLAIN IT. MAKE IT BETTER. ‘ORDER’ AND ‘SORT_BY’ - IS THERE A DIFFERENCE?
65. READ CODE. EXPLAIN IT. MAKE IT BETTER. WHY THE TERNARY?
66. READ CODE. EXPLAIN IT. MAKE IT BETTER. WHY THE ‘TO_S’? WHY THE ‘TO_I’
67. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW EXPLAINING CODE - REVIEW
68. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW EXPLAINING CODE - REVIEW ▸ What is the code doing? ▸ Learn how they translate code into human understanding
69. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW EXPLAINING CODE - REVIEW ▸ What is the code doing? ▸ Learn how they translate code into human understanding ▸ Explain the differences between ‘order’ and ‘sort_by’ ▸ Depth of knowledge of tools used on the job ▸ Do they consider tradeoffs and the potential for breaking things
70. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW EXPLAINING CODE - REVIEW ▸ What is the code doing? ▸ Learn how they translate code into human understanding ▸ Explain the differences between ‘order’ and ‘sort_by’ ▸ Depth of knowledge of tools used on the job ▸ Do they consider tradeoffs and the potential for breaking things ▸ Why the ternary? Why the ‘to_s’? ▸ Knowledge of the language’s treatment of operations on different types
71. READ CODE. EXPLAIN IT. MAKE IT BETTER. HAS THIS IMPROVED?
72. READ CODE. EXPLAIN IT. MAKE IT BETTER. HAS THIS IMPROVED? IS IT MORE READABLE?
73. READ CODE. EXPLAIN IT. MAKE IT BETTER. HAS THIS IMPROVED?
74. READ CODE. EXPLAIN IT. MAKE IT BETTER. WHAT MAKES THIS A NAIVE SOLUTION? HAS THIS IMPROVED?
75. READ CODE. EXPLAIN IT. MAKE IT BETTER. HAS THIS IMPROVED?
76. READ CODE. EXPLAIN IT. MAKE IT BETTER. ANY ISSUES WITH THIS SOLUTION? HAS THIS IMPROVED?
77. READ CODE. EXPLAIN IT. MAKE IT BETTER. HAS THIS IMPROVED?
78. READ CODE. EXPLAIN IT. MAKE IT BETTER. ANY ISSUES WITH THIS SOLUTION? HAS THIS IMPROVED?
79. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW RECAP OF FIRST EXAMPLE - DID THEY ASK GOOD QUESTIONS
80. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW RECAP OF FIRST EXAMPLE - DID THEY ASK GOOD QUESTIONS ▸ About writing tests?
81. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW RECAP OF FIRST EXAMPLE - DID THEY ASK GOOD QUESTIONS ▸ About writing tests? ▸ About the database? ▸ Foreign key constraints, Indexing, Postgres vs MySQL, etc.
82. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW RECAP OF FIRST EXAMPLE - DID THEY ASK GOOD QUESTIONS ▸ About writing tests? ▸ About the database? ▸ Foreign key constraints, Indexing, Postgres vs MySQL, etc. ▸ About Rails version and Ruby version
83. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW RECAP OF FIRST EXAMPLE - DID THEY ASK GOOD QUESTIONS ▸ About writing tests? ▸ About the database? ▸ Foreign key constraints, Indexing, Postgres vs MySQL, etc. ▸ About Rails version and Ruby version ▸ About commit messages or reasons behind the initial implementation?
84. BUT DOES IT SCALE? Jeff Bezos - Philosophy on Life Also credited to every anonymous Ruby hater, ever.
85. READ CODE. EXPLAIN IT. MAKE IT BETTER.
86. READ CODE. EXPLAIN IT. MAKE IT BETTER. WHAT ARE SOME CONCERNS WITH THIS CODE? HOW DO WE MAKE IT BETTER?
87. READ CODE. EXPLAIN IT. MAKE IT BETTER SMALLER. WHAT ABOUT A NEW OBJECT?
88. READ CODE. EXPLAIN IT. MAKE IT BETTER SMALLER.
89. MARK SIEMERS RUBYCONF 2018 CURATING YOUR OWN INTERVIEW QUESTIONS ▸ When doing a bug fix or code refactor, save a gist/snippet of before and after with notes ▸ Or search commit history ▸ Pull a ticket from your current sprint ▸ Run a churn vs complexity report or flog score REFACTORING THE TECHNICAL INTERVIEW
91. Move this way
92. WHAT IF MY CODE IS CONFIDENTIAL?
93. WHAT IF MY CODE IS CONFIDENTIAL? USE OPEN SOURCE!
94. MARK SIEMERS RUBYCONF 2018 REFACTORING THE TECHNICAL INTERVIEW USE OPEN SOURCE REPOSITORIES ▸ Use Open Source Repositories ▸ Ruby - 539k Repos ▸ Crystal - 908 Repos ▸ C - 197k Repos ▸ Python - 517k Repos ▸ C# - 173k Repos ▸ Elixir - 3k Repos ▸ JavaScript - 749k Repos ▸ Java - 753k Repos ▸ Objective-J - 443 Repos
95. THANK YOU Mark Siemers Refactoring the Coding Interview mark@measurabl.com