Beyond the Hype: Practical lessons in Long-Term Rails maintenance

New Rails features like Hotwire, Kamal, and Solid AllTheThings are super cool, but not everyone uses them daily. Many of us work on Rails apps that have grown large and complex. To keep these apps thriving, we need to stay up to date, deal with technical debt, and add new features that are important for our business needs while ensuring nothing breaks for our customers.

In this talk, I’ll give you a real-life look at how a small team of engineers keeps an almost 20-year-old monolithic Rails application at Harvest running strong and ready for the future. I’ll share how our workflows have changed (like how we write PR descriptions or how we advocate for refactoring through exploration), how we stay up to date with Rails updates (including why we chose to utilize new out-of-the-box technologies and why we decided to keep older libraries in other situations), and the practical lessons and strategies we use to keep things running smoothly, like observability, feature flags, and other utils and gems like scientist.

Transcript

1. Beyond the Hype Julia López – Balkan Ruby 2025 Practical lessons in Long-Term Rails maintenance
2. Julia López From Barcelona ☀ ❤ Rails since 2011 Refactoring 🧹 & Upgrades ⏫ 🔗 https://julialopez.dev
5. Danny Wen - https://blog.groovehq.com/harvest-danny-wen-interview “We’ve been using Excel for timesheets for our clients, and we thought, it would be so much easier if this was on the web.”
7. What is this talk about? 🤷
8. The “Hype” Why “Beyond the Hype”? ⭐ ⭐ ⭐ ⭐ ⭐ ⭐ ⭐ ⭐ ⭐ ⭐ ⭐
9. ActiveStorage
10. kt-paperclip https://github.com/kreeti/kt-paperclip
11. ActiveJob
12. Sidekiq Workers https://sidekiq.org/
13. On the other side…
14. seamless_database_pool https://github.com/bdurand/seamless_database_pool
15. AR native multiple database support https://www.rubyevents.org/talks/using-multiple-databases-with-active-record
16. tubestock https://github.com/ngauthier/tubesock
17. ActionCable No talk (yet?) 😬
18. Maintenance For the long term ⚙
19. https://dictionary.cambridge.org/dictionary/english/maintenance “the work needed to keep a road, building, machine, etc. in good condition"
20. Keeping the app running 🏃
21. Implementing new features ➕
22. Fixing user-facing bugs 🐛
23. Patching CVE 🧯
24. Dealing with last-minute deprecations 💣
25. But also…
26. Refactoring 🧽
27. Deleting dead code 🧟
28. Proactive management of dependencies 🔁
29. Enhancing the test suite ✔
30. Taking care of CI 👷 💸
31. Research 🕵
32. Monitoring 👀
33. Improving performance ⚡
34. Sharing knowledge 👨🏫
35. Knowing when to let go… 🥲
36. It me
37. Why is important?
45. Technical debt is inevitable
47. How do we do maintenance? How does maintenance look across all phases of development
48. Shared responsability
49. Boy Scout rule - https://deviq.com/principles/boy-scout-rule “Leave your code better than you found it.”
50. Project management
51. • Thoughtful estimations based on research and/or how old the code is • Advocate for research, improvements, and cleanups! • Help de f ine and shape features Project management
52. Project management • Thoughtful estimations based on research and/or how old the code is • Advocate for research, improvements, and cleanups! • Help de f ine and shape features
53. • Thoughtful estimations based on research and/or how old the code is • Advocate for research, improvements, and cleanups! • Help de f ine and shape features Project management
55. Implementation
56. • Proactive update of dependencies used in the code you are currently working on • Use latest patterns when implementing new features • Try to improve or delete the code you see on your way Implementation
57. • Proactive update of dependencies used in the code you are currently working on • Use latest patterns when implementing new features • Try to improve or delete the code you see on your way Implementation
58. Implementation • Proactive update of dependencies used in the code you are currently working on • Use latest patterns when implementing new features • Try to improve or delete the code you see on your way!
59. Pull requests & code reviews
60. • Verbose PR descriptions • Enforced a minimum number of reviewers • Advocate for maintainability on your review • Everybody is welcome to chime in! Pull requests & code reviews
61. Many of us at some point “These are complicated changes we discussed. More [LINK THAT DOESN’T LONGER WORK]”
62. • Verbose PR descriptions • Enforce a minimum number of reviewers • Advocate for maintainability on your review • Everybody is welcome to chime in! Pull requests & code reviews
63. • Verbose PR descriptions • Enforce a minimum number of reviewers • Advocate for maintainability on your review • Everybody is welcome to chime in! Pull requests & code reviews
64. • Verbose PR descriptions • Enforce a minimum number of reviewers • Advocate for maintainability on your review • Everybody is welcome to chime in! Pull requests & code reviews
65. Quality Assurance
66. • Not a safety net • They prevent issues, not just catch them • Check beyond whether the feature works or not Quality Assurance
67. • Not a safety net • They prevent issues, not just catch them • Check beyond whether the feature works or not Quality Assurance
68. • Not a safety net • They prevent issues, not just catch them • Check beyond whether the feature works or not Quality Assurance
69. Deployment and beyond
70. • The engineer who deploys the change is responsible for making sure they have not introduced any regressions • Follow-up on needed changes based on observations in production Deployment and beyond
71. • The engineer who deploys the change is responsible for making sure they have not introduced any regressions • Follow-up on needed changes based on observations in production Deployment and beyond
72. Is that it?
73. Delta Force
74. • Engineers rotate every sprint • Handle bugs outside of the scope of the squads • Oneshot requests • Help with troubleshooting • Low-hanging fruit improvements Delta Force
75. • Engineers rotate every sprint • Handle bugs outside of the scope of the squads • Oneshot requests • Help with troubleshooting • Low-hanging fruit improvements Delta Force
76. • Engineers rotate every sprint • Handle bugs outside of the scope of the squads • Oneshot requests • Help with troubleshooting • Low-hanging fruit improvements Delta Force
77. • Engineers rotate every sprint • Handle bugs outside of the scope of the squads • Oneshot requests • Help with troubleshooting • Low-hanging fruit improvements Delta Force
78. • Engineers rotate every sprint • Handle bugs outside of the scope of the squads • Oneshot requests • Help with troubleshooting • Low-hanging fruit improvements Delta Force
79. Platform Development
80. • Engineering productivity (local setup, linting, automation, CI/CD, etc.) • Observability • Core system functionalities • Major upgrades • Security Platform Development
81. • Engineering productivity (local setup, linting, automation, CI/CD, etc.) • Observability • Core system functionalities • Major upgrades • Security Platform Development
82. • Engineering productivity (local setup, linting, automation, CI/CD, etc.) • Observability • Core system functionalities • Major upgrades • Security Platform Development
83. • Engineering productivity (local setup, linting, automation, CI/CD, etc.) • Observability • Core system functionalities • Major upgrades • Security Platform Development
84. • Engineering productivity (local setup, linting, automation, CI/CD, etc.) • Observability • Core system functionalities • Major upgrades • Security Platform Development
85. Exploratory days
86. Engineering discussions
87. AI
88. Refactoring, full rewrites or do nothing? Choose your adventure!
89. Spikes 🗡
90. • Explore unknowns before building a feature • Time-boxed • Deliver POC, f indings, and recommendations • Allows us to plan and estimate better Spikes
91. • Explore unknowns before building a feature • Time-boxed • Deliver POC, f indings, and recommendations • Allows us to plan and estimate better Spikes
92. • Explore unknowns before building a feature • Time-boxed • Deliver POC, f indings, and recommendations • Allows us to plan and estimate better Spikes
93. • Explore unknowns before building a feature • Time-boxed • Deliver POC, f indings, and recommendations • Allows us to plan and estimate better Spikes
94. Adapt current functionality
95. Full rewrites
96. Look for other options
97. Do nothing
98. Dependencies managament Keep all things updated!
99. Before…
101. Dependabot https://docs.github.com/en/code-security/dependabot
102. version: 2 updates: - package-ecosystem: “docker" directory: ".docker/app/" schedule: interval: “monthly” # We don't want any general updates, only security ones open-pull-requests-limit: 0 - package-ecosystem: "bundler" directory: "/" schedule: interval: "daily" # We don't want any general updates, only security ones open-pull-requests-limit: 0 - package-ecosystem: "npm" directory: "/" schedule: interval: "daily" # We don't want any general updates, only security ones open-pull-requests-limit: 0
103. version: 2 updates: - package-ecosystem: “docker" directory: ".docker/app/" schedule: interval: “monthly” # We don't want any general updates, only security ones open-pull-requests-limit: 0 - package-ecosystem: "bundler" directory: "/" schedule: interval: "daily" # We don't want any general updates, only security ones open-pull-requests-limit: 0 - package-ecosystem: "npm" directory: "/" schedule: interval: "daily" # We don't want any general updates, only security ones open-pull-requests-limit: 0
104. version: 2 updates: - package-ecosystem: “docker" directory: ".docker/app/" schedule: interval: “monthly” # We don't want any general updates, only security ones open-pull-requests-limit: 0 - package-ecosystem: "bundler" directory: "/" schedule: interval: "daily" # We don't want any general updates, only security ones open-pull-requests-limit: 0 - package-ecosystem: "npm" directory: "/" schedule: interval: "daily" # We don't want any general updates, only security ones open-pull-requests-limit: 0
105. version: 2 updates: - package-ecosystem: “docker" directory: ".docker/app/" schedule: interval: “monthly” # We don't want any general updates, only security ones open-pull-requests-limit: 0 - package-ecosystem: "bundler" directory: "/" schedule: interval: "daily" # We don't want any general updates, only security ones open-pull-requests-limit: 0 - package-ecosystem: "npm" directory: "/" schedule: interval: "daily" # We don't want any general updates, only security ones open-pull-requests-limit: 0
106. Renovate https://github.com/renovatebot/renovate
107. // For gems and ruby versions, do major upgrades one at a time. For example, // go from 1.0.0 to 2.0.0 and then to 3.0.0 instead of 1.0.0 to 3.0.0. { matchManagers: ['bundler', 'ruby-version'], separateMultipleMajor: true, },
108. // For gems and ruby versions, do major upgrades one at a time. For example, // go from 1.0.0 to 2.0.0 and then to 3.0.0 instead of 1.0.0 to 3.0.0. { matchManagers: ['bundler', 'ruby-version'], separateMultipleMajor: true, },
109. // For rails, separate patch and minor PR's. For example, first go from // 6.0.4 to 6.0.8 and then to 6.1.X instead of 6.1.X directly. { matchManagers: ['bundler'], matchPackageNames: ['rails'], separateMinorPatch: true, },
110. // For rails, separate patch and minor PR's. For example, first go from // 6.0.4 to 6.0.8 and then to 6.1.X instead of 6.1.X directly. { matchManagers: ['bundler'], matchPackageNames: ['rails'], separateMinorPatch: true, },
111. // Update DEFAULT_REF_INTEGRATION_SUITE in .buildkite/pipeline.yml { customType: 'regex', description: 'Update the integration-suite ref in pipeline configurations', fileMatch: [ '^.buildkite/pipeline\\.yml$', ], matchStrings: [ '\\sDEFAULT_REF_INTEGRATION_SUITE: ["\'](?<currentDigest>.*)["\']\\s', ], branch: 'main', datasourceTemplate: 'git-refs', depNameTemplate: 'https://github.com/harvesthq/integration-suite', },
112. // Update DEFAULT_REF_INTEGRATION_SUITE in .buildkite/pipeline.yml { customType: 'regex', description: 'Update the integration-suite ref in pipeline configurations', fileMatch: [ '^.buildkite/pipeline\\.yml$', ], matchStrings: [ '\\sDEFAULT_REF_INTEGRATION_SUITE: ["\'](?<currentDigest>.*)["\']\\s', ], branch: 'main', datasourceTemplate: 'git-refs', depNameTemplate: 'https://github.com/harvesthq/integration-suite', },
113. // Update DEFAULT_REF_INTEGRATION_SUITE in .buildkite/pipeline.yml { customType: 'regex', description: 'Update the integration-suite ref in pipeline configurations', fileMatch: [ '^.buildkite/pipeline\\.yml$', ], matchStrings: [ '\\sDEFAULT_REF_INTEGRATION_SUITE: ["\'](?<currentDigest>.*)["\']\\s', ], branch: 'main', datasourceTemplate: 'git-refs', depNameTemplate: 'https://github.com/harvesthq/integration-suite', },
114. // Update DEFAULT_REF_INTEGRATION_SUITE in .buildkite/pipeline.yml { customType: 'regex', description: 'Update the integration-suite ref in pipeline configurations', fileMatch: [ '^.buildkite/pipeline\\.yml$', ], matchStrings: [ '\\sDEFAULT_REF_INTEGRATION_SUITE: ["\'](?<currentDigest>.*)["\']\\s', ], branch: 'main', datasourceTemplate: 'git-refs', depNameTemplate: 'https://github.com/harvesthq/integration-suite', },
115. // Update DEFAULT_REF_INTEGRATION_SUITE in .buildkite/pipeline.yml { customType: 'regex', description: 'Update the integration-suite ref in pipeline configurations', fileMatch: [ '^.buildkite/pipeline\\.yml$', ], matchStrings: [ '\\sDEFAULT_REF_INTEGRATION_SUITE: ["\'](?<currentDigest>.*)["\']\\s', ], branch: 'main', datasourceTemplate: 'git-refs', depNameTemplate: 'https://github.com/harvesthq/integration-suite', },
116. In the meantime…
117. bundle outdated
118. ❯ bundle outdated Fetching gem metadata from https://rubygems.org/......... Resolving dependencies... Gem Current Latest Requested Groups actioncable 7.2.2.1 8.0.2 actionmailbox 7.2.2.1 8.0.2 actionmailer 7.2.2.1 8.0.2 actionpack 7.2.2.1 8.0.2 actiontext 7.2.2.1 8.0.2 actionview 7.2.2.1 8.0.2 activejob 7.2.2.1 8.0.2 activemodel 7.2.2.1 8.0.2 activerecord 7.2.2.1 8.0.2 activestorage 7.2.2.1 8.0.2 activesupport 7.2.2.1 8.0.2 attr_encrypted 4.0.0 4.2.0 >= 0 default browser 4.0.0 6.2.0 >= 0 default prometheus-client 4.2.2 4.2.4 >= 0 default rack 2.2.13 3.1.13 ~> 2 default rack-session 1.0.2 2.1.0 rack-utf8_sanitizer 1.9.1 1.10.1 >= 0 default rackup 1.0.1 2.2.1 rails 7.2.2.1 8.0.2 = 7.2.2.1 default railties 7.2.2.1 8.0.2 rubocop 1.75.1 1.75.2 = 1.75.1 development, test sidekiq 7.3.1 8.0.2 >= 0 default sidekiq-cron 1.12.0 2.2.0 >= 0 default sidekiq-pro 7.3.1 8.0.1 >= 0 default xero-ruby 4.3.1 11.0.0 >= 0 default
119. bundle update
120. bundle update –-group test
121. bundle update –-group development
122. bundle_update_interactive https://github.com/mattbrictson/bundle_update_interactive
124. Major dependency upgrades?
125. The way to production… How do we deploy all these changes?
126. A test suite you can trust 👏
127. Observability
128. Continuous Integration and Deployment
129. Feature f lags ❓
131. app/views/layout/application.html.erb <% if @company.new_main_nav? %> <%= render "shared/new_main_nav" %> <% else %> <%= render "shared/main_nav" %> <% end %>
132. app/views/layout/application.html.erb <% if @company.new_main_nav? %> <%= render "shared/new_main_nav" %> <% else %> <%= render "shared/main_nav" %> <% end %>
133. • We don’t have to keep long-running branches • A/B testing • Toggle features on/o ff without deploying – great rollback mechanism • Roll out to speci f ic users or groups – test in production with internal companies Feature f lags
134. • We don’t have to keep long-running branches • A/B testing • Toggle features on/o ff without deploying – great rollback mechanism • Roll out to speci f ic users or groups – test in production with internal companies Feature f lags
135. • We don’t have to keep long-running branches • A/B testing • Toggle features on/o ff without deploying – great rollback mechanism • Roll out to speci f ic users or groups – test in production with internal companies Feature f lags
136. • We don’t have to keep long-running branches • A/B testing • Toggle features on/o ff without deploying – great rollback mechanism • Roll out to speci f ic users or groups – test in production with internal companies Feature f lags
137. Rollback strategy ⏮
138. YOLO deploying 🤞
139. There is something else…
140. scientist https://github.com/github/scientist 🔬
141. https://github.com/github/scientist “A Ruby library for carefully refactoring critical paths.”
143. class MyWidget def allows?(user) model.check_user(user).valid? end end How scientist works?
144. require "scientist" class MyWidget def allows?(user) experiment = Scientist::Default.new "widget-permissions" experiment.use { model.check_user(user).valid? } # old way experiment.try { user.can?(:read, model) } # new way experiment.run end end How scientist works?
145. require "scientist" class MyWidget def allows?(user) experiment = Scientist::Default.new "widget-permissions" experiment.use { model.check_user(user).valid? } # old way experiment.try { user.can?(:read, model) } # new way experiment.run end end How scientist works?
146. require "scientist" class MyWidget def allows?(user) experiment = Scientist::Default.new "widget-permissions" experiment.use { model.check_user(user).valid? } # old way experiment.try { user.can?(:read, model) } # new way experiment.run end end How scientist works?
147. require "scientist" class MyWidget def allows?(user) experiment = Scientist::Default.new "widget-permissions" experiment.use { model.check_user(user).valid? } # old way experiment.try { user.can?(:read, model) } # new way experiment.run end end How scientist works?
148. require "scientist" class MyWidget def allows?(user) experiment = Scientist::Default.new "widget-permissions" experiment.use { model.check_user(user).valid? } # old way experiment.try { user.can?(:read, model) } # new way experiment.run end end How scientist works?
149. What happens next?
150. require "scientist/experiment" class MyExperiment include Scientist::Experiment attr_accessor :name def initialize(name) @name = name end def enabled? true end def raised(operation, error) p "Operation '#{operation}' failed with error '#{error.inspect}'" super # will re-raise end def publish(result) p result end end How scientist works?
151. require "scientist/experiment" class MyExperiment include Scientist::Experiment attr_accessor :name def initialize(name) @name = name end def enabled? true end def raised(operation, error) p "Operation '#{operation}' failed with error '#{error.inspect}'" super # will re-raise end def publish(result) p result end end How scientist works?
152. require "scientist/experiment" class MyExperiment include Scientist::Experiment attr_accessor :name def initialize(name) @name = name end def enabled? true end def raised(operation, error) p "Operation '#{operation}' failed with error '#{error.inspect}'" super # will re-raise end def publish(result) p result end end How scientist works?
153. require "scientist/experiment" class MyExperiment include Scientist::Experiment attr_accessor :name def initialize(name) @name = name end def enabled? true end def raised(operation, error) p "Operation '#{operation}' failed with error '#{error.inspect}'" super # will re-raise end def publish(result) p result end end How scientist works?
154. require "scientist/experiment" class MyExperiment include Scientist::Experiment attr_accessor :name def initialize(name) @name = name end def enabled? true end def raised(operation, error) p "Operation '#{operation}' failed with error '#{error.inspect}'" super # will re-raise end def publish(result) p result end end How scientist works?
155. require "scientist/experiment" class MyExperiment include Scientist::Experiment attr_accessor :name def initialize(name) @name = name end def enabled? true end def raised(operation, error) p "Operation '#{operation}' failed with error '#{error.inspect}'" super # will re-raise end def publish(result) p result end end How scientist works?
156. ✋
157. class ApplicationExperiment include Scientist::Experiment attr_reader :name def initialize(name, enabled: false) @name = name @enabled = enabled end def enabled? @enabled end def publish(result) Prom.summaries["scientist_experiments"].observe result.control.duration, name: name, observation: "control" Prom.summaries["scientist_experiments"].observe result.candidates.first.duration, name: name, observation: "candidate" if result.matched? Prom.counters["science_experiments"].observe 1, name: name, kind: "match" else Prom.counters["science_experiments"].observe 1, name: name, kind: "mismatch" store_mismatch_data(result) end end private def store_mismatch_data(result) payload = { name: name, context: context, execution_order: result.observations.map(&:name), control: observation_payload(result.control), candidate: observation_payload(result.candidates.first), } key = "science.#{name}.mismatch" $redis.lpush key, payload.to_json $redis.ltrim key, 0, 50 end def observation_payload(observation) if observation.raised? { exception: observation.exception.class, message: observation.exception.message, backtrace: observation.exception.backtrace, } else { value: observation.cleaned_value } end end end Default experiment
158. Default experiment class ApplicationExperiment include Scientist::Experiment attr_reader :name def initialize(name, enabled: false) @name = name @enabled = enabled end def enabled? @enabled end def publish(result) Prom.summaries["scientist_experiments"].observe result.control.duration, name: name, observation: "control" Prom.summaries["scientist_experiments"].observe result.candidates.first.duration, name: name, observation: "candidate" if result.matched? Prom.counters["science_experiments"].observe 1, name: name, kind: "match" else Prom.counters["science_experiments"].observe 1, name: name, kind: "mismatch" store_mismatch_data(result) end end private def store_mismatch_data(result) payload = { name: name, context: context, execution_order: result.observations.map(&:name), control: observation_payload(result.control), candidate: observation_payload(result.candidates.first),
159. Default experiment class ApplicationExperiment include Scientist::Experiment attr_reader :name def initialize(name, enabled: false) @name = name @enabled = enabled end def enabled? @enabled end def publish(result) Prom.summaries["scientist_experiments"].observe result.control.duration, name: name, observation: "control" Prom.summaries["scientist_experiments"].observe result.candidates.first.duration, name: name, observation: "candidate" if result.matched? Prom.counters["science_experiments"].observe 1, name: name, kind: "match" else Prom.counters["science_experiments"].observe 1, name: name, kind: "mismatch" store_mismatch_data(result) end end private def store_mismatch_data(result) payload = { name: name, context: context, execution_order: result.observations.map(&:name), control: observation_payload(result.control), candidate: observation_payload(result.candidates.first),
160. attr_reader :name def initialize(name, enabled: false) @name = name @enabled = enabled end def enabled? @enabled end def publish(result) Prom.summaries["scientist_experiments"].observe result.control.duration, name: name, observation: "control" Prom.summaries["scientist_experiments"].observe result.candidates.first.duration, name: name, observation: "candidate" if result.matched? Prom.counters["science_experiments"].observe 1, name: name, kind: "match" else Prom.counters["science_experiments"].observe 1, name: name, kind: "mismatch" store_mismatch_data(result) end end private def store_mismatch_data(result) payload = { name: name, context: context, execution_order: result.observations.map(&:name), control: observation_payload(result.control), candidate: observation_payload(result.candidates.first), } key = "science.#{name}.mismatch" Default experiment
161. What’s our main use case?
162. We have A LOT of customer data
163. We allow our customers to report on that data
164. Creative uses of our application
165. But there’s more!
166. Past experiments
167. Experiment with SQL queries experiment = Scientist::Default.new.new("dtr-scope-experiment", enabled: true) experiment.use { control_scope } experiment.try { candidate_scope } experiment.compare do |control_scope, candidate_scope| control_scope.pluck(:id) == candidate_scope.pluck(:id) end experiment.clean { |scope| scope.pluck(:id) } experiment.run
168. Experiment with SQL queries experiment = Scientist::Default.new.new("dtr-scope-experiment", enabled: true) experiment.use { control_scope } experiment.try { candidate_scope } experiment.compare do |control_scope, candidate_scope| control_scope.pluck(:id) == candidate_scope.pluck(:id) end experiment.clean { |scope| scope.pluck(:id) } experiment.run
169. Experiment with SQL queries experiment = Scientist::Default.new.new("dtr-scope-experiment", enabled: true) experiment.use { control_scope } experiment.try { candidate_scope } experiment.compare do |control_scope, candidate_scope| control_scope.pluck(:id) == candidate_scope.pluck(:id) end experiment.clean { |scope| scope.pluck(:id) } experiment.run
170. Experiment with SQL queries experiment = Scientist::Default.new.new("dtr-scope-experiment", enabled: true) experiment.use { control_scope } experiment.try { candidate_scope } experiment.compare do |control_scope, candidate_scope| control_scope.pluck(:id) == candidate_scope.pluck(:id) end experiment.clean { |scope| scope.pluck(:id) } experiment.run
171. Experiment with SQL queries experiment = Scientist::Default.new.new("dtr-scope-experiment", enabled: true) experiment.use { control_scope } experiment.try { candidate_scope } experiment.compare do |control_scope, candidate_scope| control_scope.pluck(:id) == candidate_scope.pluck(:id) end experiment.clean { |scope| scope.pluck(:id) } experiment.run
172. Experiment with SQL queries experiment = Scientist::Default.new.new("dtr-scope-experiment", enabled: true) experiment.use { control_scope } experiment.try { candidate_scope } experiment.compare do |control_scope, candidate_scope| control_scope.pluck(:id) == candidate_scope.pluck(:id) end experiment.clean { |scope| scope.pluck(:id) } experiment.run ✅
173. Experiment with caching experiment = Scientist::Default.new(name: "project-tab-without-inner-caching", enabled: true) experiment.context = { project: @project } experiment.use { cache_query { scope.connection.select_all(scope).to_a } } experiment.try { scope.connection.select_all(scope).to_a } experiment.run
174. Experiment with caching experiment = Scientist::Default.new(name: "project-tab-without-inner-caching", enabled: true) experiment.context = { project: @project } experiment.use { cache_query { scope.connection.select_all(scope).to_a } } experiment.try { scope.connection.select_all(scope).to_a } experiment.run
175. Experiment with caching experiment = Scientist::Default.new(name: "project-tab-without-inner-caching", enabled: true) experiment.context = { project: @project } experiment.use { cache_query { scope.connection.select_all(scope).to_a } } experiment.try { scope.connection.select_all(scope).to_a } experiment.run
176. Experiment with caching experiment = Scientist::Default.new(name: "project-tab-without-inner-caching", enabled: true) experiment.context = { project: @project } experiment.use { cache_query { scope.connection.select_all(scope).to_a } } experiment.try { scope.connection.select_all(scope).to_a } experiment.run ✅
177. Remove custom implementation experiment = Scientist::Default.new(name: "basic-auth", enabled: true) experiment.use { http_auth_user_and_password } experiment.try { ActionController::HttpAuthentication::Basic.user_name_and_password(request) } experiment.run
178. Remove custom implementation experiment = Scientist::Default.new(name: "basic-auth", enabled: true) experiment.use { http_auth_user_and_password } experiment.try { ActionController::HttpAuthentication::Basic.user_name_and_password(request) } experiment.run
179. Remove custom implementation experiment = Scientist::Default.new(name: "basic-auth", enabled: true) experiment.use { http_auth_user_and_password } experiment.try { ActionController::HttpAuthentication::Basic.user_name_and_password(request) } experiment.run ✅
180. “Current” experiment
181. CurrentAttributes experiment experiment = Scientist::Default.new.new("current-user-experiment", enabled: true) experiment.use { @current_user.id } experiment.try { Current.user&.id } experiment.run
182. CurrentAttributes experiment experiment = Scientist::Default.new.new("current-user-experiment", enabled: true) experiment.use { @current_user.id } experiment.try { Current.user&.id } experiment.run ❓
183. Grafana
184. THANK YOU 🥰 Any questions? Wanna share experiences? Find me around! 🔗 https://julialopez.dev