Slide 1

Slide 1 text

Ahmed El-Helw June 18, 2022 Productive Mobile Development Tips and Tricks for Productive Mobile Development

Slide 2

Slide 2 text

No content

Slide 3

Slide 3 text

Mobile Apps are the connection to the customer

Slide 4

Slide 4 text

Source

Slide 5

Slide 5 text

Faster Iterations are Essential

Slide 6

Slide 6 text

Factors affecting Speed Things to Consider • Distractions

Slide 7

Slide 7 text

Source

Slide 8

Slide 8 text

Factors affecting Speed Things to Consider • Distractions • Communication Problems

Slide 9

Slide 9 text

No content

Slide 10

Slide 10 text

Factors affecting Speed Things to Consider • Distractions • Communication Problems • Con fl icting Priorities

Slide 11

Slide 11 text

No content

Slide 12

Slide 12 text

Factors affecting Speed Things to Consider • Distractions • Communication Problems • Con fl icting Priorities • Technical Debt

Slide 13

Slide 13 text

No content

Slide 14

Slide 14 text

Factors affecting Speed Things to Consider • Distractions • Communication Problems • Con fl icting Priorities • Technical Debt • Production Fires

Slide 15

Slide 15 text

Source

Slide 16

Slide 16 text

Factors affecting Speed Things to Consider • Distractions • Communication Problems • Con fl icting Priorities • Technical Debt • Production Fires • Hardware

Slide 17

Slide 17 text

Source

Slide 18

Slide 18 text

Process Bottlenecks Mobile Bottlenecks • Company release process • Application store review process • Application store rejections • Adoptions take time

Slide 19

Slide 19 text

Source

Slide 20

Slide 20 text

Experiments and Toggles

Slide 21

Slide 21 text

Choosing between Implementations

Slide 22

Slide 22 text

Deciding on a Direction

Slide 23

Slide 23 text

Ask the Customer Running Customer Studies • You are not your customer • Wireframes and discussions • Prepare • Question • Understand

Slide 24

Slide 24 text

Comparing between Variants

Slide 25

Slide 25 text

72.31% of all statistics are made up on the spot

Slide 26

Slide 26 text

Tim Campos “Data wins Arguments”

Slide 27

Slide 27 text

Customer Behavior Understanding and Monitoring Customer Behavior • Dashboards • critical business statistics • conversion funnels • Slice by application version • Alerts

Slide 28

Slide 28 text

Run Experiments

Slide 29

Slide 29 text

Source

Slide 30

Slide 30 text

Problems Happen

Slide 31

Slide 31 text

Dealing with Problems How to React to Problems • Ignore it for now • Hot Fix in Production • Release takes time • Adoption takes time • Risky

Slide 32

Slide 32 text

Source

Slide 33

Slide 33 text

What if we just turn it off?

Slide 34

Slide 34 text

Feature Toggle

Slide 35

Slide 35 text

Other Considerations Things to Keep in Mind • Be mindful about: • The e ff ect of experiments on each other • The e ff ect of feature fl ags on each other • Dead Experiments increase Technical Debt

Slide 36

Slide 36 text

Data

Slide 37

Slide 37 text

Data Some odds and ends about Data • Building trust with Customers • Compliance • App Stores • Governments

Slide 38

Slide 38 text

Policy Changes are often a Surprise

Slide 39

Slide 39 text

MetaFilter via Forbes Article “If you're not paying for it, you become the product”

Slide 40

Slide 40 text

As a company grows, so too does the value of its data

Slide 41

Slide 41 text

Structures and Methodologies

Slide 42

Slide 42 text

How to Structure Teams

Slide 43

Slide 43 text

Structuring Teams How to Structure Teams • Single full stack vertical teams • Separate splits between frontend and backend

Slide 44

Slide 44 text

It Depends

Slide 45

Slide 45 text

Evaluate and Adopt

Slide 46

Slide 46 text

You are not Spotify

Slide 47

Slide 47 text

Scrum? Agile?

Slide 48

Slide 48 text

It Depends

Slide 49

Slide 49 text

Source

Slide 50

Slide 50 text

Source

Slide 51

Slide 51 text

Communication is Key

Slide 52

Slide 52 text

RFCs

Slide 53

Slide 53 text

Structuring Code

Slide 54

Slide 54 text

Modularize

Slide 55

Slide 55 text

Define a Set of Rules

Slide 56

Slide 56 text

Split API and Implementation

Slide 57

Slide 57 text

Better Isolation of Features

Slide 58

Slide 58 text

Easier Sharing

Slide 59

Slide 59 text

Faster Build Times

Slide 60

Slide 60 text

Ability to Stub Implementations

Slide 61

Slide 61 text

Faster Testing

Slide 62

Slide 62 text

Standalone Apps

Slide 63

Slide 63 text

Sharing Code Across Platforms

Slide 64

Slide 64 text

How do we Save Time?

Slide 65

Slide 65 text

Options Sharing Code Between Platforms • Frameworks that handle Everything • Do it yourself Tooling

Slide 66

Slide 66 text

React Native

Slide 67

Slide 67 text

5 Part Series React Native at Airbnb

Slide 68

Slide 68 text

Flutter

Slide 69

Slide 69 text

Frameworks that Handle Everything Sharing Code Between Platforms • Pros • Potentially easier to hire for • Write once, use everywhere • Faster development (hot reload, etc) • Potential for hot loading code

Slide 70

Slide 70 text

Frameworks that Handle Everything Sharing Code Between Platforms • Cons • Need to explicitly separate UI for iOS and Android • Doesn’t “feel” native irrespective • New platform features are di ffi cult to adopt - need to wait • Overhead on binary size • Rejections Likely

Slide 71

Slide 71 text

C/C++

Slide 72

Slide 72 text

C/C++ Sharing Code Between Platforms • Pros • High performance • Easier to obfuscate code

Slide 73

Slide 73 text

C/C++ Sharing Code Between Platforms • Cons • Di ffi cult • Still have to handle di ff erences between platforms • Custom development environment

Slide 74

Slide 74 text

Dropbox Blog Post The (not so) hidden cost of sharing code between iOS and Android

Slide 75

Slide 75 text

Kotlin Multiplatform

Slide 76

Slide 76 text

Kotlin Multiplatform Sharing Code Between Platforms • Pros • Completely native • Easy to learn for engineers writing Swift • Lots of third party libraries • Backed by JetBrains and Google • Supports a plethora of platforms

Slide 77

Slide 77 text

Kotlin Multiplatform Sharing Code Between Platforms • Cons • iOS inter-op with Objective-C instead of Swift • Convincing iOS Engineers is Di ffi cult • Tooling • Still in the process of maturing

Slide 78

Slide 78 text

Which one?

Slide 79

Slide 79 text

It Depends

Slide 80

Slide 80 text

What to Share

Slide 81

Slide 81 text

Analytics

Slide 82

Slide 82 text

Business Logic

Slide 83

Slide 83 text

More?

Slide 84

Slide 84 text

Benefits and Costs Sharing Code Across Platforms - KMP • Pros: • Keeps iOS and Android consistent • Maintenance is cheaper • In the long run, development may become cheaper

Slide 85

Slide 85 text

Benefits and Costs Sharing Code Across Platforms - KMP • Cons: • Might be more expensive initially • Di ffi cult to get buy in from iOS engineers • Inter-op with Objective-C instead of Swift

Slide 86

Slide 86 text

What to Share Across Platforms

Slide 87

Slide 87 text

Vocabulary

Slide 88

Slide 88 text

Features

Slide 89

Slide 89 text

Architecture

Slide 90

Slide 90 text

Design System

Slide 91

Slide 91 text

Versioning

Slide 92

Slide 92 text

Release Process

Slide 93

Slide 93 text

CI/CD

Slide 94

Slide 94 text

Production Issues take Time

Slide 95

Slide 95 text

Time is Money

Slide 96

Slide 96 text

Catching Mistakes Early Saves Time and Money

Slide 97

Slide 97 text

Defense Mechanisms

Slide 98

Slide 98 text

Code Review

Slide 99

Slide 99 text

Tests Catching Issues Early • Tests • Unit Tests • Integration Tests • On-Device Integration Tests (“UI Tests”) • Screenshot Tests

Slide 100

Slide 100 text

Automated Checks

Slide 101

Slide 101 text

Lint

Slide 102

Slide 102 text

Code Styling Enforcer

Slide 103

Slide 103 text

Code Smell Detection

Slide 104

Slide 104 text

Expose Hidden Changes Catching Issues Early • Transitive Dependencies • Binary Size • Permissions

Slide 105

Slide 105 text

Release Early

Slide 106

Slide 106 text

Release Often

Slide 107

Slide 107 text

Staged Rollouts

Slide 108

Slide 108 text

Beta Program

Slide 109

Slide 109 text

Engineering Culture

Slide 110

Slide 110 text

Engineers are Humans

Slide 111

Slide 111 text

Humans make Mistakes

Slide 112

Slide 112 text

Blame is Bad

Slide 113

Slide 113 text

Blame is not Productive

Slide 114

Slide 114 text

Learning from Mistakes is Productive

Slide 115

Slide 115 text

How could we have prevented it?

Slide 116

Slide 116 text

Build Ownership

Slide 117

Slide 117 text

Hire Good Engineering Managers

Slide 118

Slide 118 text

Give Developers Good Hardware

Slide 119

Slide 119 text

Keep Learning and Improving

Slide 120

Slide 120 text

Focus on Impact

Slide 121

Slide 121 text

Thank You

Slide 122

Slide 122 text

Q & A