The World Runs on Bad Software

Transcript

1. Brandon Keepers
   Lone Star Ruby Conference V
   August 12, 2011
   BAD
   the world runs on
   software
   Ordered List
   

   View Slide

2. View Slide

3. @bkeepers
   I

   View Slide

4. opensoul.org
   I

   View Slide

5. View Slide

6. group therapy
   

   View Slide

7. View Slide

8. View Slide

9. pride
   

   View Slide

10. craftsman
    

    View Slide

11. art
    

    View Slide

12. I

    View Slide

13. get.harmonyapp.com
    

    View Slide

14. get.gaug.es
    

    View Slide

15. speakerdeck.com
    

    View Slide

16. View Slide

17. I — GOOD CODE!
    

    View Slide

18. not perfect
    

    View Slide

19. :(
    

    View Slide

20. HAS NOT
    IMPLODED
    

    View Slide

21. HAS NOT
    IMPLODED yet
    ^
    

    View Slide

22. RUNS
    

    View Slide

23. in spite of
    

    View Slide

24. because of
    

    View Slide

25. bad software
    

    View Slide

26. pragmatism
    perfection
    

    View Slide

27. What is good software?
    How does it go bad?
    How do we build better software?
    

    View Slide

28. good software
    

    View Slide

29. “
    Good software meets the
    present objective at an
    acceptable cost without
    incurring an unacceptable
    amount of future risk.
    

    View Slide

30. adequacy
    

    View Slide

31. !mediocrity
    

    View Slide

32. objective
    

    View Slide

33. software
    !=
    code
    

    View Slide

34. “
    Zach Dennis
    Mutually Human Software
    People fail to realize that
    quality is merely a property of
    the code that exists, not the
    reason for its existence.
    

    View Slide

35. users don’t care
    

    View Slide

36. “
    Scott Berkun
    We’re frustrated most in
    life by things that come
    close to our deepest
    needs, but don’t deliver.
    

    View Slide

37. cost
    

    View Slide

38. “
    Ray Yeargin
    Virtually all of the cost of
    software development is,
    directly and indirectly, the
    cost of design.
    

    View Slide

39. Production
    almost zero of
    

    View Slide

40. $$$$
    

    View Slide

41. $$$
    

    View Slide

42. $$
    

    View Slide

43. $
    

    View Slide

44. View Slide

45. risk
    

    View Slide

46. !science
    

    View Slide

47. ‑Risk
    ‐Cost
    

    View Slide

48. What is good software?
    How does it go bad?
    How do we build better software?
    

    View Slide

49. entropy
    

    View Slide

50. “
    The Pragmatic Programmer
    While software development
    is immune from almost all
    physical laws, entropy hits
    us hard.
    

    View Slide

51. why?
    

    View Slide

52. software is
    malleable
    

    View Slide

53. View Slide

54. software is
    complex
    

    View Slide

55. rake stats
    

    View Slide

56. 2.5 million parts
    

    View Slide

57. View Slide

58. software is
    immature
    

    View Slide

59. View Slide

60. software is
    opaque
    

    View Slide

61. software is
    durable
    

    View Slide

62. View Slide

63. What is good software?
    How does it go bad?
    How do we build better software?
    

    View Slide

64. better software?
    how do we build
    

    View Slide

65. talk to your users
    

    View Slide

66. automated testing
    

    View Slide

67. insurance policy
    

    View Slide

68. automation
    computers are good at
    

    View Slide

69. View Slide

70. DRY
    

    View Slide

71. View Slide

72. refactoring
    small, incremental
    

    View Slide

73. What is good software?
    How does it go bad?
    How do we build better software?
    

    View Slide

74. Examples
    

    View Slide

75. method_missing
    

    View Slide

76. class NilClass
    def method_missing(method, *args)
    nil
    end
    end
    

    View Slide

77. >> nil.foo.bar
    => nil
    >> nil + 5
    => nil
    >> 5 + nil
    TypeError: coerce must return [x, y]
    

    View Slide

78. View Slide

79. performance
    maintenance
    

    View Slide

80. def to_xml(options={})
    root_tag = options.fetch(:root, 'game')
    instruct = options[:skip_instruct] ? '' : XML_INSTRUCT
    # …
    # gather up a bunch of variables
    # …
    <<-EOF
    #{instruct}
    <#{root_tag}>
    #{id}
    #{current_move_user_id}
    current-move-user-id>
    #{created_by_user_id}
    created-by-user-id>
    0
    #{is_matchmaking}
    #{was_matchmaking}matchmaking>
    #{moves_count}
    #{random_seed}
    

    View Slide

81. #{instruct}
    <#{root_tag}>
    #{id}
    #{current_move_user_id}
    current-move-user-id>
    #{created_by_user_id}
    created-by-user-id>
    0
    #{is_matchmaking}
    #{was_matchmaking}matchmaking>
    #{moves_count}
    #{random_seed}
    #{client_version}
    #{created_at.try(:xmlschema)}
    created-at>
    #{users_xml}
    #{current_user_xml}
    #{moves_xml}
    #{chats_xml}
    #{root_tag}>
    EOF
    end
    

    View Slide

82. commit fa779d9c2d1214e43ff497f386b51ebf3708346e
    Author: Brandon Keepers
    Date: Wed Sep 22 17:08:38 2010 -0400
    First attempt at getting embedded callbacks working
    

    View Slide

83. def create_or_update
    run_callbacks(:before_save)
    if result = super
    run_callbacks(:after_save)
    end
    result
    end
    Rails 2
    

    View Slide

84. def create_or_update
    run_callbacks(:save) { super }
    end
    Rails 3
    

    View Slide

85. def run_callbacks(kind, options={}, &block)
    self.embedded_associations.each do |association|
    self.send(association.name).run_callbacks(
    kind, options, &block
    )
    end
    end
    

    View Slide

86. def run_callbacks(callback, &block)
    embedded_docs = []
    embedded_associations.each do |association|
    embedded_docs += association.target
    end
    block = embedded_docs.inject(block) do |chain, doc|
    if doc.class.respond_to?("_#{callback}_callbacks")
    lambda { doc.run_callbacks(callback, &chain) }
    else
    chain
    end
    end
    super callback, &block
    end
    

    View Slide

87. SystemStackError: stack level too deep
    from mongo_mapper/plugins/embedded_callbacks.rb:23:in `run_callbacks'
    from activesupport-3.0.9/lib/active_support/callbacks.rb:414:in
    `_run_create_callbacks'
    from activesupport-3.0.9/lib/active_support/callbacks.rb:94:in `send'
    from activesupport-3.0.9/lib/active_support/callbacks.rb:94:in
    `run_callbacks'
    from mongo_mapper/plugins/embedded_callbacks.rb:28:in `run_callbacks'
    from mongo_mapper/plugins/embedded_callbacks.rb:23:in `run_callbacks'
    

    View Slide

88. now what?
    

    View Slide

89. pragmatic
    

    View Slide

90. software is design
    

    View Slide

91. embrace constraints
    

    View Slide

92. engineering is design
    

    View Slide

93. craftsmanship is design
    

    View Slide

94. “
    Glen Vanderburg
    Sometimes you have to move
    ahead with weak evidence and
    weak conclusions because you
    don't have time to prove them.
    

    View Slide

95. opportunity costs
    

    View Slide

96. be pragmatic
    

    View Slide

97. View Slide

98. “
    Ed Yourdon
    You can discipline yourself to write
    software that's good enough—good
    enough for your users, for future
    maintainers, for your own peace of
    mind. You'll find that you are more
    productive and your users are happier.
    

    View Slide

99. Ordered List
    Thank you!
    brandon@orderedlist.com
    Brandon Keepers
    Lone Star Ruby Conference V
    August 12, 2011
    @bkeepers
    http://spkr8.com/t/7829
    

    View Slide

100. Credits
     Power lines - http://www.flickr.com/photos/viamoi/3339707547/
     House of cards - http://www.flickr.com/photos/gibbons/2294374741/
     Dentists office - http://www.flickr.com/photos/cmdrcord/4996672560/
     Broken chair - http://www.flickr.com/photos/gorbould/3004789912/
     Schematics - http://www.flickr.com/photos/nicklockey/4547935355/
     Engineers - http://www.flickr.com/photos/seattlemunicipalarchives/4818952324/
     Black box - http://goldberg.berkeley.edu/art/big-images/
     

     View Slide