How to plan a feature (long version)

Transcript

1. Radoslav Stankov How to Plan a Feature
3. Radoslav Stankov @rstankov
9. Complicated feature !
10. Where to start?
11. Waterfall methodology
12. " Waterfall Methodology
13. ... this what we get
15. Agile methodology
16. # Agile Methodology
18. No need for upfront design, just...
19. ... code like a cowboy
21. ... this what we get
26. Tools when you need them...
29. Plan for today
30. $ Drawing % Spike & Focused Task ' Feature Switches ( Agenda
31. $ Drawing
33. ) Draw state/flow we want to achieve * Draw current state/flow + Database structure ➡ What we need todo (plan) $ Drawing
34. - Tools Excalidraw - great for single player Whimsical - great for collaboration
 
 PlantUML - great for documentation Figma - for UI collaboration with designer Miro/FigJam - for group brainstorming ... use whatever works for you .
35. https://blog.rstankov.com/how-i-plan-and-execute-features/
37. done todo won't do in progress https://blog.rstankov.com/how-i-plan-and-execute-features/
40. / Showcase 1
41. Feature: Stacking on product page product page (not in stack) -> do you use -> yes -> review
 product page (not in stack) -> do you use -> no -> search (alternative) -> review 
 product page (in stack) -> still using? -> yes -> ask for tip -> add tip
 product page (in stack) -> still using? -> no -> search (alternative) -> review / Showcase 1 (end to end test scenarios)
42. / Showcase 2
43. / Showcase 2
44. / Showcase 3 (brainstorm)
45. / Showcase 4 (advice)
46. / Showcase 5 (advice)
47. PlantUML https://www.plantuml.com/
48. @startuml "Browser" -> "Cloudflare": HTTPS Request "Cloudflare" -> "AWS ELB": HTTPS Request "AWS ELB" -> "Frontend (AWS ECS)": HTTPS Request "Frontend (AWS ECS)" -> "Backend (AWS ECS)": GraphQL Request "Backend (AWS ECS)" -> "Redis (Rediscloud)": Cache Query "Backend (AWS ECS)" -> "Database (AWS RDS Aurora)": SQL Query "Backend (AWS ECS)" --> "Frontend (AWS ECS)": JSON Response "Frontend (AWS ECS)" --> "Browser": HTML Response @enduml PlantUML https://www.plantuml.com/
51. ) Explain the problem to someone (vocal or in writing) * Why something won't work 0 While doing this ideas will come 0 Explain the problem
52. ) Create empty file on Desktop * Write quick & dirty code in empty file + Imagine a perfect world for this code 1 Delete file # If solution is 2 ➡ think of something else ➡ If solution is ✅ ➡ use ideas in project ✍ Code Sketch
53. % Spike
54. “A task aimed at answering a question or gathering information, rather than at producing shippable product. Sometimes we can't estimate until the development team does some actual work to resolve a technical question or a design problem.” - WikipediA Spike (software development)
55. “A task aimed at answering a question or gathering information, rather than at producing shippable product. Sometimes we can't estimate until the development team does some actual work to resolve a technical question or a design problem.” - WikipediA Spike (software development)
56. 5 Define scope and time-box ) Write quick & dirty code in a Pull Request * Review the solution + Close Pull Request without merging # If solution is 2 ➡ Discuss new solution ➡ If solution is ✅ ➡ Open new Pull Request % Spike
58. “The goal of a "spike" is to help you draw a map of how to implement a solution with confidence, not to implement a solution.” - Rado
59. “The best feature of a "spike" is that it is a safe way to say this approach doesn't work.” - Rado
61. 6 ...but “We can't do a "spike" for every feature. It will waste too much time.” “You don't need a "spike" to show that your solution isn't good.” “Why do you need to throw away the code?” “Isn't this a Prototype, MVP, PoC?”
62. 6 ...but “We can't do a "spike" for every feature. It will waste too much time.” “You don't need a "spike" to show that your solution isn't good.” “Why do you need to throw away the code?” “Isn't this a Prototype, MVP, PoC?” Don't do it for every feature. Do it when you are not confident how to solve a problem. 7
63. 6 ...but “We can't do a "spike" for every feature. It will waste too much time.” “You don't need a "spike" to show that your solution isn't good.” “Why do you need to throw away the code?” “Isn't this a Prototype, MVP, PoC?” Because of "Tunnel vision" and "sunk cost fallacy" 7
64. 6 ...but “We can't do a "spike" for every feature. It will waste too much time.” “You don't need a "spike" to show that your solution isn't good.” “Why do you need to throw away the code?” “Isn't this a Prototype, MVP, PoC?” Because it is usually a mess. I don't delete it, I just copy parts I need into a clean new PR 7
65. 6 ...but “We can't do a "spike" for every feature. It will waste too much time.” “You don't need a "spike" to show that your solution isn't good.” “Why do you need to throw away the code?” “Isn't this a Prototype, MVP, PoC?” Kinda - yes. The difference is in how big it is, who is meant for team or external, and do we ship it in production. .
66. 6 ...but “We can't do a "spike" for every feature. It will waste too much time.” “You don't need a "spike" to show that your solution isn't good.” “Why do you need to throw away the code?” “Isn't this a Prototype, MVP, PoC?”
67. & Focused Task
69. What's next 8
70. Focused Task
71. Focused Task https://github.com/RStankov/FocusedTask
72. Focused Task is just sketchpad for my current task. For everything else I use "Todoist"
73. ' Feature switches
77. OFF ON User ' Feature switches Check
79. 1 week 2 week 3 week ...
80. 1 week 2 week 3 week ... developer
81. 1 week 2 week 3 week ... developer beta
82. 1 week 2 week 3 week ... developer beta admin
83. 1 week 2 week 3 week ... developer beta admin everybody
84. Project Iteration Delivery Retro
85. unless Feature.enabled?(:funds) redirect_to feature_unavailable_path
 end 9 Ruby on Rails
86. <Feature isEnabled="fund">
 <Link to={paths.funds}>Your funds</Link>
 </Feature> ⚛ React
87. const fundsEnabled = useFeature("funds") ⚛ React
90. Recap
92. ... of course all depends on specific feature .
94. Thanks ;
95. https://speakerdeck.com/rstankov