ceej's how to solve it

Maxims applicable to problem-solving in engineering organizations and in all other areas of life.

Transcript

1. ceej's how to solve it
2. ceej's how to solve it, or engineering at npm
3. a series of maxims & their application to problem-solving of diverse kinds
5. dedicated to David Zink with whom I have spent 20 happy years arguing about all this
6. it takes decades to get good at writing software
7. most human beings learn complex skills through mentorship
8. most programmers learn by working with more experienced programmers
9. our industry's booms have caused shortages of experienced mentors
10. exacerbated by a modern contempt for anybody over 40
11. too often programmers come of age without experienced people around them
12. we keep re-inventing core tenets of our profession
13. npm's engineering staff meeting me attempting to share what I know and let you share what you know
14. lift each other up
15. therefore... this talk
17. systems are how things work together
18. the npm's registry is a system
19. npm itself is a system!
20. systems analyst once the title for people who do what programmers do now
21. systems exist in a constant state of failure The Systems Bible by John Gall
22. "Every complex working system can be shown to have evolved from a simple working system"
23. from this we extract a maxim...
24. write the simplest thing you can, every time
25. evolve your simple systems to do more keep them working the whole time
26. this is true for people systems as well as for computers & code
27. Simplicity is prerequisite for reliability. —Edsger W. Dijkstra
28. simple is not the same as easy Rich Hickey: Simple Made Easy
29. Do the simplest thing that moves you toward your goal while preserving your flexibility. —ceej
31. modularity
32. modularity is one of the big secrets
33. modularity is hiding the details of an implementation
34. if you hide the details you can change it later
35. the other secret is that change is guaranteed so prepare for it
36. the easy connection is with what npm does: we manage a javascript modules
37. modularity's more important applications are at a systems level
38. we hide our data storage implementations behind microservices
39. nobody knows if we're using mysql, postgres, or mongodb we can change it without other services caring
40. just did this to the downloads API mysql ➜ redis
42. always pronounce the names of things in the funniest possible way
43. don't let the perfect be the enemy of the better than what we have now
44. continuous improvement our products, our processes, our systems
45. iterate toward victory!
46. don't be clever
47. don't be clever “Debugging is twice as hard as writing the code in the first place. Therefore, if you write the code as cleverly as possible, you are, by definition, not smart enough to debug it. ” —Brian Kernighan
49. think
50. the RFC process is about stopping to think
51. what problem are you solving?
52. Defining the problem clearly helps you understand what a solution looks like.
53. RFC stands for request for comment
54. seek feedback
56. people make the best decisions they can given their information & abilities
57. improving the quality of your information will improve your decisions
58. give good feedback in return
59. your colleagues rely on you just as you rely on them
60. tie your ego to being right in the end not to being right from the start
62. you can't fool me, young man. it's tradeoffs all the way down
63. what's a tradeoff?
64. you accept something that's meh in one way but that's great in a way that matters more
65. every solution to a problem requires tradeoffs
66. make tradeoffs with your eyes open
68. I'll phrase that more positively
69. be kind be courageous
70. be kind to each other
71. kindness doesn't mean never saying hard things
72. kindness shapes how you say them
73. what does courage have to do with problem-solving?
74. be bold about tackling hard problems
75. be bold about re-examining what we do now
76. sometimes you need to blow everything up
78. the companies that survive are phoenixes
79. all code has a lifespan
80. code is originally written to solve a specific problem
81. our understanding of the problem changes the context around it changes the code itself changes
82. plan to replace all solutions to problems
83. feel good about how long your solution lasted!
85. how we deal with mistakes is important
86. because mistakes are inevitable
87. we try for blameless postmortems
88. we redesign processes when things go wrong
89. the right thing to do should be the easiest thing to do
90. work hard to make the right thing easy
91. do small things daily to tidy up
94. all code is communication
95. all of our work requires communication
96. learning our professions requires communication
97. how much time in each day do you spend talking to your colleagues?
98. do the work to get better at communicating
99. find where you can get better & do it! writing? chatting in slack? listening to feedback? giving feedback?
100. communication skills always pay off
102. never be cruel or cowardly if you are always make amends -- The Twelfth Doctor