Code That Fits in Your Head

Praise for
Code That Fits in Your Head

“We progress in software by standing on the shoulders of those who came
before us. Mark’s vast experience ranges from philosophical and organisa-
tional considerations right down to the precise details of writing code. In this
book, you’re offered an opportunity to build on that experience. Use it.”
—Adam Ralph, speaker, tutor, and software simplifier, Particular Software
“I’ve been reading Mark’s blogs for years and he always manages to entertain
while at the same time offering deep technical insights. Code That Fits in Your
Head follows in that vein, offering a wealth of information to any software
developer looking to take their skills to the next level.”
—Adam Tornhill, founder of CodeScene, author of Software Design
X-Rays and Your Code as a Crime Scene
“My favorite thing about this book is how it uses a single code base as a
working example. Rather than having to download separate code samples, you
get a single Git repository with the entire application. Its history is hand-
crafted to show the evolution of the code alongside the concepts being
explained in the book. As you read about a particular principle or technique,
you’ll find a direct reference to the commit that demonstrates it in practice. Of
course, you’re also free to navigate the history at your own leisure, stopping at
any stage to inspect, debug, or even experiment with the code. I’ve never seen
this level of interactivity in a book before, and it brings me special joy because
it takes advantage of Git’s unique design in a new constructive way.”
—Enrico Campidoglio, independent consultant, speaker and
Pluralsight author
“Mark Seemann not only has decades of experience architecting and building
large software systems, but is also one of the foremost thinkers on how to scale
and manage the complex relationship between such systems and the teams
that build them.”
—Mike Hadlow, freelance software consultant and blogger
“Mark Seemann is well known for explaining complex concepts clearly and
thoroughly. In this book he condenses his wide-ranging software development
experience into a set of practical, pragmatic techniques for writing sustainable
and human-friendly code. This book will be a must read for every
programmer.”
—Scott Wlaschin, author of Domain Modeling Made Functional
“Mark writes, ‘Successful software endures’—this book will help you to write
that kind of software.”
—Bryan Hogan, software architect, podcaster, blogger
“Mark has an extraordinary ability to help others think deeply about the
industry and profession of software development. With every interview on
.NET Rocks! I have come away knowing I would have to go back and listen to
my own show to really take in everything we discussed.”
—Richard Campbell, co-host, .NET Rocks!
Code That Fits
in Your Head
Code That Fits
in Your Head
Heuristics for Software Engineering
Mark Seemann
Boston • Columbus • New York • San Francisco • Amsterdam • Cape Town

Dubai • London • Madrid • Milan • Munich • Paris • Montreal • Toronto • Delhi • Mexico City
São Paulo • Sydney • Hong Kong • Seoul • Singapore • Taipei • Tokyo
Cover: Mark Seeman
Page xxix, author photo: © Linea Vega Seemann Jacobsen
Page 12, Queen Alexandrine’s Bridge, Denmark: Ulla Seemann
Page 33, baseball and bat: buriy/123RF
Page 38, illustration of human brain: maglyvi/Shutterstock
Page 38, illustration of laptop computer: grmarc/Shutterstock
Page 157, Figure 8.2: © Microsoft 2021
Page 158, Figure 8.3, scissors: Hurst Photo/Shutterstock
Page 158, Figure 8.3, hand saw: Andrei Kuzmik/Shutterstock
Page 158, Figure 8.3, utility knife: Yogamreet/Shutterstock
Page 158, Figure 8.3, Phillips-head screwdriver: bozmp/Shutterstock
Page 158, Figure 8.3, Swiss military knife: Billion Photos/Shutterstock
Page 159, Figure 8.4: Roman Babakin/Shutterstock
Page 170, Figure 8.5: © Microsoft 2021
Page 239, Figure 12.1: ajt/Shutterstock
Page 259, Figure 13.2, bursting star: Arcady/Shutterstock
Page 269, Figure 13.5: Verdandi/123RF
Page 277, Figure 14.1: Tatyana Pronina/Shutterstock
Page 291, Figure 15.2: kornilov007/Shutterstock
Page 291, hammer: bozmp/Shutterstock
Pages 306, Figure 15.3: Figure based on a screen shot from codescene.io
Pages 307, Figure 15.4: Figure based on a screen shot from codescene.io
Many of the designations used by manufacturers and sellers to distinguish their products are claimed as
trademarks. Where those designations appear in this book, and the publisher was aware of a trademark claim,
the designations have been printed with initial capital letters or in all capitals.
The author and publisher have taken care in the preparation of this book, but make no expressed or implied
warranty of any kind and assume no responsibility for errors or omissions. No liability is assumed for incidental
or consequential damages in connection with or arising out of the use of the information or programs contained
herein.
For information about buying this title in bulk quantities, or for special sales opportunities (which may include
electronic versions; custom cover designs; and content particular to your business, training goals, marketing
focus, or branding interests), please contact our corporate sales department at [email protected] or
(800) 382-3419.
For government sales inquiries, please contact [email protected].
For questions about sales outside the U.S., please contact [email protected].
Visit us on the Web: informit.com/aw
Library of Congress Control Number: 2021944424
Copyright © 2022 Pearson Education, Inc.
All rights reserved. This publication is protected by copyright, and permission must be obtained from
the publisher prior to any prohibited reproduction, storage in a retrieval system, or transmission in any form or
by any means, electronic, mechanical, photocopying, recording, or likewise. For information regarding
permissions, request forms and the appropriate contacts within the Pearson Education Global Rights &
Permissions Department, please visit www.pearson.com/permissions.
ISBN-13: 978-0-13-746440-1
ISBN-10: 0-13-746440-1
ScoutAutomatedPrintCode
To my parents:
My mother, Ulla Seemann, to whom I owe my attention to detail.
My father, Leif Seemann, from whom I inherited my contrarian streak.
This page intentionally left blank
“The future is already here — it’s just not very evenly distributed”
—William Gibson
C on t en t s
Series Editor Foreword xix

Preface xxiii
About the Author xxix
PART I Acceleration 1
Chapter 1 Art or Science? 3

1.1 Building a House 4
1.1.1 The Problem with Projects 4
1.1.2 The Problem with Phases 5
1.1.3 Dependencies 6
1.2 Growing a Garden 7
1.2.1 What Makes a Garden Grow? 7
1.3 Towards Engineering 8
1.3.1 Software as a Craft 8
1.3.2 Heuristics 10
1.3.3 Earlier Notions of Software Engineering 11
1.3.4 Moving Forward with Software Engineering 12
1.4 Conclusion 14
xi
Contents
Chapter 2 Checklists 15
2.1 An Aid to Memory 15
2.2 Checklist for a New Code Base 17
2.2.1 Use Git 18
2.2.2 Automate the Build 19
2.2.3 Turn On all Error Messages 24
2.3 Adding Checks to Existing Code Bases 29
2.3.1 Gradual Improvement 30
2.3.2 Hack Your Organisation 31
2.4 Conclusion 32
Chapter 3 Tackling Complexity 33

3.1 Purpose 34
3.1.1 Sustainability 35
3.1.2 Value 36
3.2 Why Programming Is Difficult 38
3.2.1 The Brain Metaphor 38
3.2.2 Code Is Read More Than It’s Written 39
3.2.3 Readability 40
3.2.4 Intellectual Work 41
3.3 Towards Software Engineering 44
3.3.1 Relationship to Computer Science 44
3.3.2 Humane Code 45
3.4 Conclusion 46
Chapter 4 Vertical Slice 49

4.1 Start with Working Software 50
4.1.1 From Data Ingress to Data Persistence 50
4.1.2 Minimal Vertical Slice 51
4.2 Walking Skeleton 53
4.2.1 Characterisation Test 54
4.2.2 Arrange Act Assert 56
4.2.3 Moderation of Static Analysis 57
4.3 Outside-in 60
4.3.1 Receive JSON 61
4.3.2 Post a Reservation 64
4.3.3 Unit Test 68
xii
Contents
4.3.4 DTO and Domain Model 70

4.3.5 Fake Object 73
4.3.6 Repository Interface 74
4.3.7 Create in Repository 74
4.3.8 Configure Dependencies 76
4.4 Complete the Slice 77
4.4.1 Schema 78
4.4.2 SQL Repository 79
4.4.3 Configuration with Database 81
4.4.4 Perform a Smoke Test 82
4.4.5 Boundary Test with Fake Database 83
4.5 Conclusion 85
Chapter 5 Encapsulation 87
5.1 Save the Data 87
5.1.1 The Transformation Priority Premise 88
5.1.2 Parametrised Test 89
5.1.3 Copy DTO to Domain Model 91
5.2 Validation 92
5.2.1 Bad Dates 93
5.2.2 Red Green Refactor 96
5.2.3 Natural Numbers 99
5.2.4 Postel’s Law 102
5.3 Protection of Invariants 105
5.3.1 Always Valid 106
5.4 Conclusion 108
Chapter 6 Triangulation 111

6.1 Short-Term versus Long-Term Memory 111
6.1.1 Legacy Code and Memory 113
6.2 Capacity 114
6.2.1 Overbooking 115
6.2.2 The Devil’s Advocate 119
6.2.3 Existing Reservations 121
6.2.4 Devil’s Advocate versus Red Green Refactor 123
6.2.5 When Do You Have Enough Tests? 126
6.3 Conclusion 127
xiii
Contents
Chapter 7 Decomposition 129

7.1 Code Rot 129
7.1.1 Thresholds 130
7.1.2 Cyclomatic Complexity 132
7.1.3 The 80/24 Rule 134
7.2 Code That Fits in Your Brain 136
7.2.1 Hex Flower 136
7.2.2 Cohesion 138
7.2.3 Feature Envy 142
7.2.4 Lost in Translation 144
7.2.5 Parse, Don’t Validate 145
7.2.6 Fractal Architecture 148
7.2.7 Count the Variables 153
7.3 Conclusion 153
Chapter 8 API Design 155

8.1 Principles of API Design 156
8.1.1 Affordance 156
8.1.2 Poka-Yoke 158
8.1.3 Write for Readers 160
8.1.4 Favour Well-Named Code over Comments 160
8.1.5 X Out Names 161
8.1.6 Command Query Separation 164
8.1.7 Hierarchy of Communication 167
8.2 API Design Example 168
8.2.1 Maître D’ 169
8.2.2 Interacting with an Encapsulated Object 171
8.2.3 Implementation Details 174
8.3 Conclusion 176
Chapter 9 Teamwork 177

9.1 Git 178
9.1.1 Commit Messages 178
9.1.2 Continuous Integration 182
9.1.3 Small Commits 184
9.2 Collective Code Ownership 187
xiv
Contents
9.2.1 Pair Programming 189

9.2.2 Mob Programming 191
9.2.3 Code Review Latency 192
9.2.4 Rejecting a Change Set 194
9.2.5 Code Reviews 195
9.2.6 Pull Requests 197
9.3 Conclusion 199
PART II Sustainability 201
Chapter 10 Augmenting Code 203

10.1 Feature Flags 204
10.1.1 Calendar Flag 204
10.2 The Strangler Pattern 209
10.2.1 Method-Level Strangler 211
10.2.2 Class-Level Strangler 215
10.3 Versioning 218
10.3.1 Advance Warning 219
10.4 Conclusion 220
Chapter 11 Editing Unit Tests 223

11.1 Refactoring Unit Tests 223
11.1.1 Changing the Safety Net 224
11.1.2 Adding New Test Code 225
11.1.3 Separate Refactoring of Test and Production Code 227
11.2 See Tests Fail 233
11.3 Conclusion 234
Chapter 12 Troubleshooting 235

12.1 Understanding 235
12.1.1 Scientific Method 236
12.1.2 Simplify 237
12.1.3 Rubber Ducking 238
12.2 Defects 240
12.2.1 Reproduce Defects as Tests 241
12.2.2 Slow Tests 243
12.2.3 Non-deterministic Defects 246
xv
Contents
12.3 Bisection 250

12.3.1 Bisection with Git 251
12.4 Conclusion 255
Chapter 13 Separation of Concerns 257

13.1 Composition 258
13.1.1 Nested Composition 258
13.1.2 Sequential Composition 262
13.1.3 Referential Transparency 264
13.2 Cross-Cutting Concerns 267
13.2.1 Logging 267
13.2.2 Decorator 268
13.2.3 What to Log 272
13.3 Conclusion 274
Chapter 14 Rhythm 275

14.1 Personal Rhythm 276
14.1.1 Time-Boxing 276
14.1.2 Take Breaks 278
14.1.3 Use Time Deliberately 279
14.1.4 Touch Type 280
14.2 Team Rhythm 282
14.2.1 Regularly Update Dependencies 282
14.2.2 Schedule Other Things 283
14.2.3 Conway’s Law 284
14.3 Conclusion 285
Chapter 15 The Usual Suspects 287

15.1 Performance 288
15.1.1 Legacy 288
15.1.2 Legibility 290
15.2 Security 292
15.2.1 STRIDE 292
15.2.2 Spoofing 294
15.2.3 Tampering 294
15.2.4 Repudiation 296
xvi
Contents
15.2.5 Information Disclosure 296

15.2.6 Denial of Service 298
15.2.7 Elevation of Privilege 299
15.3 Other Techniques 300
15.3.1 Property-Based Testing 300
15.3.2 Behavioural Code Analysis 305
15.4 Conclusion 308
Chapter 16 Tour 309

16.1 Navigation 309
16.1.1 Seeing the Big Picture 310
16.1.2 File Organisation 314
16.1.3 Finding Details 316
16.2 Architecture 318
16.2.1 Monolith 318
16.2.2 Cycles 319
16.3 Usage 323
16.3.1 Learning from Tests 323
16.3.2 Listen to Your Tests 325
16.4 Conclusion 326
Appendix A List of Practices 329

A.1 The 50/72 Rule 329
A.2 The 80/24 Rule 330
A.3 Arrange Act Assert 330
A.4 Bisection 330
A.5 Checklist for A New Code Base 331
A.6 Command Query Separation 331
A.7 Count the Variables 331
A.8 Cyclomatic Complexity 331
A.9 Decorators for Cross-Cutting Concerns 332
A.10 Devil’s Advocate 332
A.11 Feature Flag 332
A.12 Functional Core, Imperative Shell 333
A.13 Hierarchy of Communication 333
A.14 Justify Exceptions from the Rule 333
xvii
Contents
A.15 Parse, Don’t Validate 334

A.16 Postel’s Law 334
A.17 Red Green Refactor 334
A.18 Regularly Update Dependencies 335
A.19 Reproduce Defects as Tests 335
A.20 Review Code 335
A.21 Semantic Versioning 335
A.22 Separate Refactoring of Test and Production Code 335
A.23 Slice 336
A.24 Strangler 336
A.25 Threat-Model 337
A.26 Transformation Priority Premise 337
A.27 X-driven Development 337
A.28 X Out Names 338
Bibliography 339
Index 349
xviii
Series Ed itor
Fo r ewor d
My grandson is learning to code.
Yes, you read that right. My 18-year-old grandson is learning to program

computers. Who’s teaching him? His aunt, my youngest daughter, who was
born in 1986, and who 16 months ago decided to change careers from chemical
engineering to programming. And who do they both work for? My eldest son,
who along with my youngest son, is in the process of starting up his second
software consultancy.
Yeah, software runs in the family. And, yeah, I’ve been programming for a
long, long time.
Anyway, my daughter asked me to spend an hour with my grandson teaching

him about the basics and the beginnings of computer programming. So we
started up a Tuple session and I lectured him on what computers were, and
how they got started, and what early computers looked like, and . . . well,
you know.
By the end of the lecture I was coding up the algorithm for multiplying two
binary integers, in PDP-8 assembly language. For those of you who aren’t
aware, the PDP-8 had no multiply instruction; you had to write an algorithm
xix
Series Editor Foreword
to multiply numbers. Indeed, the PDP-8 didn’t even have a subtract

instruction; you had to use two’s complement and add a pseudo-negative
number (let the reader understand).
As I finished up the coding example, it occurred to me that I was scaring my

grandson to death. I mean, when I was 18 this kind of geeky detail thrilled me;
but maybe it wasn’t so attractive to an 18-year-old whose aunt is trying to
teach him how to write simple Clojure programs.
Anyway, it made me think of just how hard programming actually is. And it
is hard. It’s really hard. It may be the hardest thing that humans have ever
attempted.
Oh, I don’t mean it’s hard to write the code to calculate a bunch of prime
numbers, or a Fibonacci sequence, or a simple bubble sort. That’s not too
hard. But an Air Traffic Control system? A luggage management system? A bill
of materials system? Angry Birds? Now that’s hard. That’s really, really hard.
I’ve known Mark Seemann for quite a few years now. I don’t remember ever
actually meeting him. It may be that we have never actually been together in
the same room. But he and I have interacted quite a bit in professional
newsgroups and social networks. He’s one of my favourite people to disagree
with.
He and I disagree on all kinds of things. We disagree on static versus dynamic

typing. We disagree on operating systems and languages. We disagree on, well,
lots of intellectually challenging things. But disagreeing with Mark is
something you have to do very carefully because the logic of his arguments
is impeccable.
So when I saw this book, I thought about how much fun it was going to be to
read through and disagree with. And that’s exactly what happened. I read
through it. I disagreed with some things. And I had fun trying to find a way to
make my logic supersede his. I think I may have even succeeded in one or two
cases—in my head—maybe.
xx
Series Editor Foreword
But that’s not the point. The point is that software is hard; and much of the
last seven decades have been spent trying to find ways to make it a little bit
easier. What Mark has done in this book is to gather all the best ideas from
those seven decades and compile them in one place.
More than that, he has organized them into a set of heuristics and techniques,
and placed them in the order that you would execute them. Those heuristics
and techniques build on each other, helping you move from stage to stage
while developing a software project.
In fact, Mark develops a software project throughout the pages of this book,
while explaining each stage and the heuristics and techniques that benefit
that stage.
Mark uses C# (one of the things I disagree with ;-), but that’s not relevant.
The code is simple, and the heuristics and techniques are applicable to any
other language you might be using.
He covers things such as Checklists, TDD, Command Query Separation, Git,

Cyclomatic Complexity, Referential Transparency, Vertical Slicing, Legacy
Strangulation, and Outside-In Development, just to mention a few.
Moreover, there are gems scattered literally everywhere throughout these

pages. I mean, you’ll be reading along, and all of a sudden he’ll say something
like, “Rotate your test function 90 degrees and see if you can balance it on the
Act of the Arrange/Act/Assert triplet” or “The goal is not to write code fast.
The goal is sustainable software” or “Commit database schema to git”.
Some of these gems are profound, some are just idle mentions, others are
speculations, but all of them are examples of the deep insight that Mark has
acquired over the years.
So read this book. Read it carefully. Think through Mark’s impeccable logic.
Internalise these heuristics and techniques. Stop and consider the insightful
gems as they pop out at you. And just maybe, when it comes time for you to
lecture your grandchildren, you won’t scare the devil out of them.
—Robert C. Martin
xxi
P r eface
In the second half of the 2000s, I began doing technical reviews for a publisher.
After reviewing a handful of books, the editor contacted me about a book on
Dependency Injection.
The overture was a little odd. Usually, when they contacted me about a book,
it would already have an author and a table of contents. This time, however,
there was none of that. The editor just requested a phone call to discuss
whether the book’s subject matter was viable.
I thought about it for a few days and found the topic inspiring. At the same
time, I couldn’t see the need for an entire book. After all, the knowledge was
out there: blog posts, library documentation, magazine articles, even a few
books all touched on related topics.
On reflection, I realised that, while the information was all out there, it was
scattered, and used inconsistent and sometimes conflicting terminology.
There’d be value in collecting that knowledge and presenting it in a consistent
pattern language.
Two years later, I was the proud author of a published book.
xxiii
Preface
After some years had gone by, I began to think about writing another book.
Not this one, but a book about some other topic. Then I had a third idea, and
a fourth, but not this one.
A decade went by, and I began to realise that when I consulted teams on
writing better code, I’d suggest practices that I’d learned from better minds
than mine. And again, I realised that most of that knowledge is already
available, but it’s scattered, and few people have explicitly connected the dots
into a coherent description of how to develop software.
Based on my experience with the first book, I know that there’s value in
collecting disparate information and presenting it in a consistent way. This
book is my attempt at creating such a package.
Wh o S h o u l d R e a d T h i s B o o k
This book is aimed at programmers with at least a few years of professional
experience. I expect readers to have suffered through a few bad software
development projects; to have experience with unmaintainable code. I also
expect readers seeking to improve.
The core audience is ‘enterprise developers’—particularly back-end

developers. I’ve spent most of my career in that realm, so this simply reflects
my own expertise. But if you’re a front-end developer, a games programmer, a
development tools engineer, or something else entirely, I expect you will still
gain a lot from reading this book.
You should be comfortable reading code in a compiled, object-oriented

language in the C family. While I’ve been a C# programmer for most of my
career, I’ve learned a lot from books with example code in C++ or Java1 . This
book turns the tables: Its example code is in C#, but I hope that Java,
TypeScript, or C++ developers find it useful, too.
1. If you’re curious about which books I mean, take a look at the bibliography.
xxiv
Preface
Prerequ isites
This isn’t a beginner’s book. While it deals with how to organise and structure
source code, it doesn’t cover the most basic details. I expect that you already
understand why indentation is helpful, why long methods are problematic,
that global variables are bad, and so on. I don’t expect you to have read Code
Complete [65], but I assume that you know of some of the basics covered there.
A Note for S oftwa re Architects

The term ‘architect’ means different things to different people, even when the
context is constrained to software development. Some architects focus on the
big picture; they help an entire organisation succeed with its endeavours.
Other architects are deep in the code and mainly concerned with the
sustainability of a particular code base.
To the degree that I’m a software architect, I’m the latter kind. My expertise is
in how to organise source code so that it addresses long-term business goals. I
write about what I know, so to the degree this book is useful to architects, it
will be that type of architect.
You’ll find no content about Architecture Tradeoff Analysis Method (ATAM),

Failure Mode and Effects Analysis (FMEA), service discovery, and so on. That
kind of architecture is outside the scope of this book.
O rg a n i s at i o n
While this is a book about methodologies, I’ve structured it around a code
example that runs throughout the book. I decided to do it that way in order to
make the reading experience more compelling than a typical ‘pattern
catalogue’. One consequence of this decision is that I introduce practices and
heuristics when they fit the ‘narrative’. This is also the order in which I
typically introduce the techniques when I coach teams.
The narrative is structured around a sample code base that implements a

restaurant reservation system. The source code for that sample code base is
available at informit.com/title/9780137464401.
xxv
Preface
If you want to use the book as a handbook, I’ve included an appendix with
a list of all the practices and information about where in the book you can
read more.
Abou t th e Code Sty l e

The example code is written in C#, which is a language that has rapidly
evolved in recent years. It’s picking up more and more syntax ideas from
functional programming; as an example, immutable record types were released
while I was writing the book. I’ve decided to ignore some of these new
language features.
Once upon a time, Java code looked a lot like C# code. Modern C# code, on
the other hand, doesn’t look much like Java.
I want the code to be comprehensible to as many readers as possible. Just as

I’ve learned much from books with Java examples, I want readers to be able to
use this book without knowing the latest C# syntax. Thus, I’m trying to stick
to a conservative subset of C# that ought to be legible to other programmers.
This doesn’t change the concepts presented in the book. Yes, in some
instances, a more succinct C#-specific alternative is possible, but that would
just imply that extra improvements are available.
To Var or Not to Va r
The var keyword was introduced to C# in 2007. It enables you to declare a
variable without explicitly stating its type. Instead, the compiler infers the type
from the context. To be clear, variables declared with var are exactly as
statically typed as variables declared with explicit types.
For a long time the use of this keyword was controversial, but most people
now use it; I do, too, but I occasionally encounter pockets of resistance.
While I use var professionally, writing code for a book is a slightly different
context. Under normal circumstances, an IDE isn’t far away. A modern
development environment can quickly tell you the type of an implicitly typed
variable, but a book can’t.
xxvi
Preface
I have, for that reason, occasionally chosen to explicitly type variables. Most
of the example code still uses the var keyword because it makes the code
shorter, and line width is limited in a printed book. In a few cases, though, I’ve
deliberately chosen to explicitly declare a variable’s type, in the hope that it
makes the code easier to understand when read in a book.
Code L istin gs
The majority of the code listings are taken from the same sample code base.
It’s a Git repository, and the code examples are taken from various stages of
development. Each such code listing includes a relative path to the file in
question. Part of that file path is a Git commit ID.
For example, listing 2.1 includes this relative path: Restaurant/f729ed9/

Restaurant.RestApi/Program.cs. This means that the example is taken from
commit ID f729ed9, and the file is Restaurant.RestApi/Program.cs. In
other words, to view this particular version of the file, you check out that
commit:
$ git checkout f729ed9
When you’ve done that, you can now explore the Restaurant.RestApi/
Program.cs file in its full, executable context.
A Not e o n t h e Bi b l io g r a p h y
The bibliography contains a mix of resources, including books, blog posts,
and video recordings. Many of my sources are online, so I have of course
supplied URLs. I’ve made an effort to mostly include resources that I have
reason to believe have a stable presence on the Internet.
Still, things change. If you’re reading this book in the future, and a URL has
become invalid, try an internet archive service. As I’m writing this, https://
archive.org is the best candidate, but that site could also be gone in the future.
Q u otin g Myself
Apart from other resources, the bibliography also includes a list of my own
work. I’m aware that, as far as making a case, quoting myself doesn’t
constitute a valid argument in itself.
xxvii
Preface
I’m not including my own work as a sleight of hand. Rather, I’m including
these resources for the reader who might be interested in more details. When
I cite myself, I do it because you may find an expanded argument, or a more
detailed code example, in the resource I point to.
Ac k now l e d g e me n t s
I’d like to thank my wife Cecilie for love and support during all the years we’ve
been together, and my children Linea and Jarl for staying out of trouble.
Apart from family, my first thanks go to my invaluable long-time friend

Karsten Strøbæk, who not only has tolerated my existence for 25 years, but
who was also the first reviewer on this book. He also helped me with various
LATEX tips and tricks, and added more entries to the index than I did.
I’d also like to thank Adam Tornhill for his feedback on the section about his
work.
I’m indebted to Dan North for planting the phrase Code That Fits in Your
Head in my subconscious, which might have happened as early as 2011 [72].
Register your copy of Code That Fits in Your Head on the InformIT site
for convenient access to updates and/or corrections as they become
available. To start the registration process, go to informit.com/register
and log in or create an account. Enter the product ISBN
(9780137464401) and click Submit. Look on the Registered Products tab
for an Access Bonus Content link next to this product, and follow that
link to access any available bonus materials. If you would like to be
notified of exclusive offers on new editions and updates, please check the
box to receive email from us.
xxviii
Ab o ut t he
Aut ho r
Mark Seemann is a bad economist who’s

found a second career as a programmer,
and he has worked as a web and enterprise
developer since the late 1990s. As a young
man Mark wanted to become a rock star, but
unfortunately had neither the talent nor the
looks – later, however, he became a Certified
Rockstar Developer. He has also written a
Jolt Award-winning book about Dependency
Injection, given more than a 100 international
conference talks, and authored video courses
for both Pluralsight and Clean Coders. He
has regularly published blog posts since 2006.
He lives in Copenhagen with his wife and two
children.
xxix
T rou b l es h oot in g
12
Professional software development consists of more than feature development.
There are also meetings, time reports, compliance activities, and ... defects.
You run into errors and problems all the time. Your code doesn’t compile, the
software doesn’t do what it’s supposed to, it runs too slowly, et cetera.
The better you get at solving problems, the more productive you are. Most of
your troubleshooting skills may be based on “the shifting sands of individual
experience” [4], but there are techniques that you can apply.
This chapter presents some of them.
12.1 U n d e r s ta n d i n g
The best advice I can think of is this:
Try to understand what’s going on.
235
Chapter 12 Troubleshooting
If you don’t understand why something doesn’t work1 , then make

understanding it a priority. I’ve witnessed a fair amount of ‘programming by
coincidence’ [50]: throw enough code at the wall to see what sticks. When it
looks as though the code works, developers move on to the next task. Either
they don’t understand why the code works, or they may fail to understand that
it doesn’t, really.
If you understand the code from the beginning, chances are that it’ll be easier
to troubleshoot.
12.1.1 S cien tific Method

When a problem manifests, most people jump straight into troubleshooting
mode. They want to address the problem. For people who program by
coincidence [50], addressing a problem typically involves trying various
incantations that may have worked before on a similar problem. If the first
magic spell doesn’t work, they move on to the next. This can involve restarting
a service, rebooting a computer, running a tool with elevated privileges,
changing small pieces of code, calling poorly-understood routines, etc. When
it looks like the problem has disappeared, they call it a day without trying to
understand why [50].
Needless to say, this isn’t an effective way to deal with problems.
Your first reaction to a problem should be to understand why it’s happening.

If you have absolutely no idea, ask for help. Usually, though, you already have
some inclination of what the problem may be. In that case, adopt a variation
of the scientific method [82]:
• Make a prediction. This is called a hypothesis.

• Perform the experiment.
• Compare outcome to prediction. Repeat until you understand what’s going
on.
1. Or, if you don’t understand why something does work.
236
12.1 Understanding
Don’t be intimidated by the term ‘scientific method’. You don’t have to don a
lab coat or design a randomised controlled double-blind trial. But do try to
come up with a falsifiable hypothesis. This might simply be a prediction, such
as “if I reboot the machine, the problem goes away,” or “if I call this function,
the return value will be 42.”
The difference between this technique and ‘programming by coincidence’ is

that the goal of going through these motions isn’t to address the problem. The
goal is to understand it.
A typical experiment could be a unit test, with a hypothesis that if you run it,
it’ll fail. See subsection 12.2.1 for more details.
12.1.2 S im p lify
Consider if removing some code can make a problem go away.
The most common reaction to a problem is to add more code to address it.
The unspoken line of reasoning seems to be that the system ‘works’, and the
problem is just an aberration. Thus, the reasoning goes, if the problem is a
special case, it should be solved with more code to handle that special case.
This may occasionally be the case, but it’s more likely that the problem is a
manifestation of an underlying implementation error. You’d be surprised how
often you can solve problems by simplifying the code.
I’ve seen plenty of examples of such an ‘action bias’ in our industry. People
who solve problems I never have because I work hard to keep my code simple:
• People develop complex Dependency Injection Containers [25] instead of

just composing object graphs in code.
• People develop complicated ‘mock object libraries’ instead of writing
mostly pure functions.
• People create elaborate package restore schemes instead of just checking
dependencies into source control.
237
• People use advanced diff tools instead of merging more frequently.

• People use convoluted object-relational mappers (ORMs) instead of
learning (and maintaining) a bit of SQL.
I could go on.
To be fair, coming up with a simpler solution is hard. For example, it took me

a decade of erecting increasingly more elaborate contraptions in
object-oriented code before I found simpler solutions. It turns out that many
things that are difficult in traditional object-oriented programming are simple
in functional programming. Once I learned about some of these concepts, I
found ways to use them in object-oriented contexts, too.
The point is that a catchphrase like KISS2 is useless in itself, because how does
one keep things simple?
You often have to be smart to keep it simple3 , but look for simplicity anyway.
Consider if there’s a way you can solve the problem by deleting code.
12.1.3 R u bber D u ck i ng
Before we discuss some specific problem-solving practices, I want to share
some general techniques. It’s not unusual to be stuck on a problem. How do
you get unstuck?
You may be staring at a problem with no clue as to how to proceed. As the

above advice goes, your first priority should be to understand the problem.
What do you do if you’re drawing a blank?
If you don’t manage your time, you can be stuck with a problem for a long
time, so do manage your time. Time-box the process. For example, set aside 25
minutes to look at the problem. If, after the time is up, you’ve made no
progress, take a break.
2. Keep It Simple, Stupid.

3. Rich Hickey discusses simplicity in Simple Made Easy [45]. I owe much of my perspective on simplicity to
that talk.
238
12.1 Understanding
When you take a break, physically remove yourself from the computer.
Go get a cup of coffee. Something happens in your brain when you get
out of your chair and away from the screen. After a couple of minutes away
from the problem, you’ll likely begin to think about something else. Perhaps
you meet a colleague as you’re moving about. Perhaps you discover that the
coffee machine needs a refill. Whatever it is, it temporarily takes your mind off
the problem. That’s often enough to give you a fresh perspective.
I’ve lost count of the number of times I return to a problem after a stroll, only
to realise that I’ve been thinking about it the wrong way.
If walking about for a few minutes isn’t enough, try asking for help. If you have
a colleague to bother, do that.
I’ve experienced this often enough: I start explaining the problem, but halfway
in, I break off in mid-sentence: “Never mind, I’ve just gotten an idea!”
The mere act of explaining a problem tends to produce new insight.
If you don’t have a colleague, you may try explaining the problem to a rubber
duck, such as the one shown in figure 12.1.
Figure 12.1 A rubber duck. Talk to it. It’ll solve your problems.
It doesn’t really have to be a rubber duck, but the technique is known as

rubber ducking because one programmer actually did use one [50].
239
Instead of using a rubber duck, I typically begin writing a question on the

Stack Overflow Q&A site. More often than not, I realise what the problem is
before I’m done formulating the question4 .
And if realisation doesn’t come, I have a written question that I can publish.
12.2 D e f e c t s
I once started in a new job in a small software startup. I soon asked my
co-workers if they’d like to use test-driven development. They hadn’t used it
before, but they were keen on learning new things. After I’d shown them the
ropes, they decided that they liked it.
A few months after we’d adopted test-driven development, the CEO came by
to talk to me. He mentioned in passing that he’d noticed that since we’d
started using tests, defects in the wild had significantly dropped.
That still makes me proud to this day. The shift in quality was so dramatic that
the CEO had noticed. Not by running numbers or doing a complex analysis,
but simply because it was so significant that it called attention to itself.
You can reduce the number of defects, but you can’t eliminate them. But do
yourself a favour: don’t let them accumulate.
The ideal number of defects is zero.
Zero bugs isn’t as unrealistic as it sounds. In lean software development, this is

known as building quality in [82]. Don’t push defects in front of you to ‘deal
with them later’. In software development, later is never.
4. When that happens, I don’t succumb to the sunk cost fallacy. Even if I’ve spent time writing the question,
I usually delete it because I deem that it’s not, after all, of general interest.
240
12.2 Defects
When a bug appears, make it a priority to address it. Stop what you’re doing5
and fix the defect instead.
12.2.1 R ep rodu ce Def ects a s Tes ts

Initially, you may not even understand what the problem is, but when you think
that you do, perform an experiment: The understanding should enable you to
formulate a hypothesis, which again enables you to design an experiment.
Such an experiment may be an automated test. The hypothesis is that when

you run the test, it’ll fail. When you actually do run the test, if it does fail,
you’ve validated the hypothesis. As a bonus, you also have a failing test that
reproduces the defect, and that will later serve as a regression test.
If, on the other hand, the test succeeds, the experiment failed. This means that
your hypothesis was wrong. You’ll need to revise it so that you can design a
new experiment. You may need to repeat this process more than once.
When you finally have a failing test, ‘all’ you have to do is to make it pass. This
can occasionally be difficult, but in my experience, it usually isn’t. The hard
part of addressing a defect is understanding and reproducing it.
I’ll show you an example from the online restaurant reservation system. While
I was doing some exploratory testing I noticed something odd when I updated
a reservation. Listing 12.1 shows an example of the issue. Can you spot the
problem?
The problem is that the email property holds the name, and vice versa. It
seems that I accidentally switched them around somewhere. That’s the initial
hypothesis, but it may take a little investigation to figure out where.
Have I not been following test-driven development? Then how could this
happen?
5. Isn’t it wonderful that with Git you can simply stash your current work?
241
Listing 12.1 Updating a reservation with a PUT request. A defect is manifest in this interaction.
Can you spot it?
PUT /reservations/21b4fa1975064414bee402bbe09090ec HTTP/1.1
Content-Type: application/json
{
"at": "2022-03-02 19:45",
"email": "[email protected]",
"name": "Phil Anders",
"quantity": 2
}
HTTP/1.1 200 OK
Content-Type: application/json; charset=utf-8
{
"id": "21b4fa1975064414bee402bbe09090ec",
"at": "2022-03-02T19:45:00.0000000",
"email": "Phil Anders",
"name": "[email protected]",
"quantity": 2
}
This could happen because I’d implemented SqlReservationsRepository6 as

a Humble Object [66]. This is an object so simple that you may decide not to
test it. I often use the rule of thumb that if the cyclomatic complexity is 1, a
test (also with a cyclomatic complexity of 1) may not be warranted.
Even so, you can still make mistakes even when the cyclomatic complexity is 1.
Listing 12.2 shows the offending code. Can you spot the problem?
Given that you already know what the problem is, you can probably guess that
the Reservation constructor expects the email argument before the name.
Since both parameters are declared as string, though, the compiler doesn’t
complain if you accidentally swap them. This is another example of stringly
typed code [3], which we should avoid7 .
6. See for example listing 4.19.

7. One way to avoid stringly typed code is to introduce Email and Name classes that wrap their respective
string values. This prevents some cases of accidentally swapping these two arguments, but as it turned
out when I did it, it wasn’t entirely foolproof. You can consult the example code’s Git repository if you’re
interested in the details. The bottom line was that I felt that an integration test was warranted.
242
12.2 Defects
Listing 12.2 The offending code fragment that causes the defect shown in listing 12.1. Can you
spot the programmer error?
(Restaurant/d7b74f1/Restaurant.RestApi/SqlReservationsRepository.cs)
using var rdr =
await cmd.ExecuteReaderAsync().ConfigureAwait(false);
if (!rdr.Read())
return null;
return new Reservation(

id,
(DateTime)rdr["At"],
(string)rdr["Name"],
(string)rdr["Email"],
(int)rdr["Quantity"]);
It’s easy enough to address the defect, but if I can make the mistake once, I can
make it again. Thus, I want to prevent a regression. Before fixing the code, write
a failing test that reproduces the bug. Listing 12.3 shows the test I wrote. It’s an
integration test that verifies that if you update a reservation in the database and
subsequently read it, you should receive a reservation equal to the one you
saved. That’s a reasonable expectation, and it reproduces the error because the
ReadReservation method swaps name and email, as shown in listing 12.2.
That PutAndReadRoundTrip test is an integration test that involves the

database. This is new. So far in this book, all tests have been running without
external dependencies. Involving the database is worth a detour.
12.2.2 Slow Tests

Bridging the gap between a programming language’s perspective on data and a
relational database is error-prone8 , so why not test such code?
In this subsection, you’ll see an outline of how to do that, but there’s a

problem: such tests tend to be slow. They tend to be orders of magnitudes
slower than in-process tests.
8. Proponents of object-relational mappers (ORMs) might argue that this makes the case for such a tool.
As I’ve stated elsewhere in this book, I consider ORMs a waste of time: they create more problems than
they solve. If you disagree, then feel free to skip this subsection.
243
Listing 12.3 Integration test of SqlReservationsRepository.

(Restaurant/645186b/Restaurant.RestApi.SqlIntegrationTests/SqlReservationsRepositoryTests.cs)
[Theory]
[InlineData("2032-01-01 01:12", "[email protected]", "z", "Zet", 4)]
[InlineData("2084-04-21 23:21", "[email protected]", "q", "Quu", 9)]
public async Task PutAndReadRoundTrip(
string date,
string email,
string name,
string newName,
int quantity)
{
var r = new Reservation(
Guid.NewGuid(),
DateTime.Parse(date, CultureInfo.InvariantCulture),
new Email(email),
new Name(name),
quantity);
var connectionString = ConnectionStrings.Reservations;
var sut = new SqlReservationsRepository(connectionString);
await sut.Create(r);
var expected = r.WithName(new Name(newName));

await sut.Update(expected);
var actual = await sut.ReadReservation(expected.Id);
Assert.Equal(expected, actual);
}
The time it takes to execute a test suite matters, particularly for developer tests
that you continually run. When you refactor with the test suite as a safety net,
it doesn’t work if it takes half an hour to run all tests. When you follow the
Red Green Refactor process for test-driven development, it doesn’t work if
running the tests takes five minutes.
The maximum time for such a test suite should be ten seconds. If it’s much
longer than that, you’ll lose focus. You’ll be tempted to look at your email,
Twitter, or Facebook while the tests run.
You can easily eat into such a ten-second budget if you involve a database.
Therefore, move such tests to a second stage of tests. There are many ways you
can do this, but a pragmatic way is to simply create a second Visual Studio
244
12.2 Defects
solution to exist side-by-side with the day-to-day solution. When you do that,
remember to also update the build script to run this new solution instead, as
shown in listing 12.4.
Listing 12.4 Build script running all tests. The Build.sln file contains both unit and integration
tests that use the database. Compare with listing 4.2. (Restaurant/645186b/build.sh)
#!/usr/bin/env bash
dotnet test Build.sln --configuration Release
The Build.sln file contains the production code, the unit test code, as well as
integration tests that use the database. I do day-to-day work that doesn’t
involve the database in another Visual Studio solution called Restaurant.sln.
That solution only contains the production code and the unit tests, so running
all tests in that context is much faster.
The test in listing 12.3 is part of the integration test code, so only runs when I
run the build script, or if I explicitly choose to work in the Build.sln solution
instead of in Restaurant.sln. It’s sometimes practical to do that, if I need to
perform a refactoring that involves the database code.
I don’t want to go into too much detail about how the test in listing 12.3
works, because it’s specific to how .NET interacts with SQL Server. If you’re
interested in the details, they’re all available in the accompanying example
code base, but briefly, all the integration tests are adorned with a
[UseDatabase] attribute. This is a custom attribute that hooks into the
xUnit.net unit testing framework to run some code before and after each test
case. Thus, each test case is surrounded with behaviour like this:
1. Create a new database and run all DDL9 scripts against it.
2. Run the test.
3. Tear down the database.
9. Data Definition Language, typically a subset of SQL. See listing 4.18 for an example.
245
Yes, each test creates a new database only to delete it again some milliseconds
later10 . That is slow, which is why you don’t want such tests to run all the time.
Defer slow tests to a second stage of your build pipeline. You can do it as
outlined above, or by defining new steps that only run on your Continuous
Integration server.
12.2.3 Non -determ inis tic Def ects

After running the restaurant reservation system for some time, the restaurant’s
maître d’ files a bug: once in a while, the system seems to allow overbooking.
She can’t deliberately reproduce the problem, but the state of the reservations
database can’t be denied. Some days contain more reservations than the
business logic shown in listing 12.5 allows. What’s going on?
You peruse the application logs11 and finally figure it out. Overbooking is a
possible race condition. If a day is approaching capacity and two reservations
arrive simultaneously, the ReadReservations method might return the same
set of rows to both threads, indicating that a reservation is possible. As figure
12.2 shows, each thread determines that it can accept the reservation, so it
adds a new row to the table of reservations.
This is clearly a defect, so you should reproduce it with a test. The problem is,
however, that this behaviour isn’t deterministic. Automated tests are supposed
to be deterministic, aren’t they?
It is, indeed, best if tests are deterministic, but do entertain, for a moment, the
notion that nondeterminism may be acceptable. In which way could this be?
10. Whenever I explain this approach to integration testing with a database, I’m invariably met with the
reaction that one can, instead, test by rolling back transactions. Yes, except that this means that you
can’t test database transaction behaviour. Also, using transaction rollback may be faster, but have you
measured? I have, once, and found no significant difference. See also section 15.1 for my general position
on performance optimisation.
11. See subsection 13.2.1.
246
12.2 Defects
Tests can fail in two ways: A test may indicate a failure where none is;
this is called a false positive. A test may also fail to indicate an actual error;
this is called a false negative.
Listing 12.5 Apparently, there’s a bug in this code that allows overbooking. What could be the
problem? (Restaurant/dd05589/Restaurant.RestApi/ReservationsController.cs)
[HttpPost]
public async Task<ActionResult> Post(ReservationDto dto)
{
if (dto is null)
throw new ArgumentNullException(nameof(dto));
var id = dto.ParseId() ?? Guid.NewGuid();

Reservation? r = dto.Validate(id);
if (r is null)
return new BadRequestResult();
var reservations = await Repository

.ReadReservations(r.At)
.ConfigureAwait(false);
if (!MaitreD.WillAccept(DateTime.Now, reservations, r))
return NoTables500InternalServerError();
await Repository.Create(r).ConfigureAwait(false);
await PostOffice.EmailReservationCreated(r).ConfigureAwait(false);
return Reservation201Created(r);
}
Read
Reservations reservations Create OK
THREAD 1 TIME
DATABASE
THREAD 2
Read reservations Create OK
Reservations
Figure 12.2 A race condition between two threads (e.g. two HTTP clients) concurrently trying
to make a reservation.
False positives are problematic because they introduce noise, and thereby
decrease the signal-to-noise ratio of the test suite. If you have a test suite that
often fails for no apparent reason, you stop paying attention to it [31].
247
False negatives aren’t quite as bad. Too many false negatives could decrease
your trust in a test suite, but they introduce no noise. Thus, at least, you know
that if a test suite is failing, there is a problem.
One way to deal with the race condition in the reservation system, then, is to
reproduce it as the non-deterministic test in listing 12.6.
Listing 12.6 Non-deterministic test that reproduces a race condition.

(Restaurant/98ab6b5/Restaurant.RestApi.SqlIntegrationTests/ConcurrencyTests.cs)
[Fact]
public async Task NoOverbookingRace()
{
var start = DateTimeOffset.UtcNow;
var timeOut = TimeSpan.FromSeconds(30);
var i = 0;
while (DateTimeOffset.UtcNow - start < timeOut)
await PostTwoConcurrentLiminalReservations(
start.DateTime.AddDays(++i));
}
This test method is only an orchestrator of the actual unit test. It keeps
running the PostTwoConcurrentLiminalReservations method in listing 12.7
for 30 seconds, over and over again, to see if it fails. The assumption, or hope,
is that if it can run for 30 seconds without failing, the system may actually
have the correct behaviour.
There’s no guarantee that this is the case. If the race condition is as scarce as
hen’s teeth, this test could produce a false negative. That’s not my experience,
though.
When I wrote this test, it only ran for a few seconds before failing. That gives
me some confidence that the 30-second timeout is a sufficiently safe margin,
but I admit that I’m guessing; it’s another example of the art of software
engineering.
It turned out that the system had the same bug when updating existing
reservations (as opposed to creating new ones), so I also wrote a similar test
for that case.
248
12.2 Defects
Listing 12.7 The actual test method orchestrated by the code in listing 12.6. It attempts to
post two concurrent reservations. The state of the system is that it’s almost sold out (the
capacity of the restaurant is ten, but nine seats are already reserved), so only one of those
reservations should be accepted.
(Restaurant/98ab6b5/Restaurant.RestApi.SqlIntegrationTests/ConcurrencyTests.cs)
private static async Task PostTwoConcurrentLiminalReservations(
DateTime date)
{
date = date.Date.AddHours(18.5);
using var service = new RestaurantService();
var initialResp =
await service.PostReservation(new ReservationDtoBuilder()
.WithDate(date)
.WithQuantity(9)
.Build());
initialResp.EnsureSuccessStatusCode();
var task1 = service.PostReservation(new ReservationDtoBuilder()

.WithDate(date)
.WithQuantity(1)
.Build());
var task2 = service.PostReservation(new ReservationDtoBuilder()
.WithDate(date)
.WithQuantity(1)
.Build());
var actual = await Task.WhenAll(task1, task2);
Assert.Single(actual, msg => msg.IsSuccessStatusCode);

Assert.Single(
actual,
msg => msg.StatusCode == HttpStatusCode.InternalServerError);
}
These tests are examples of slow tests that ought to be included only as
second-stage tests as discussed in subsection 12.2.2.
There are various ways you can address a defect like the one discussed here.
You can reach for the Unit of Work [33] design pattern. You can also deal with
the issue at the architectural level, by introducing a durable queue with a
single-threaded writer that consumes the messages from it. In any case, you
need to serialise the reads and the writes involved in the operation.
I chose to go for a pragmatic solution: use .NET’s lightweight transactions, as

shown in listing 12.8. Surrounding the critical part of the Post method with a
249
TransactionScope effectively serialises12 the reads and writes. That solves the
problem.
Listing 12.8 The critical part of the Post method is now surrounded with a
TransactionScope, which serialises the read and write methods. The highlighted code is new
compared to listing 12.5. (Restaurant/98ab6b5/Restaurant.RestApi/ReservationsController.cs)
using var scope = new TransactionScope(
TransactionScopeAsyncFlowOption.Enabled);
var reservations = await Repository
.ReadReservations(r.At)
.ConfigureAwait(false);
if (!MaitreD.WillAccept(DateTime.Now, reservations, r))
return NoTables500InternalServerError();
await Repository.Create(r).ConfigureAwait(false);
await PostOffice.EmailReservationCreated(r).ConfigureAwait(false);
scope.Complete();
In my experience, most defects can be reproduced as deterministic tests, but

there’s a residual that eludes this ideal. Multithreaded code infamously falls
into that category. Of two evils, I prefer nondeterministic tests over no test
coverage at all. Such tests will often have to run until they time out in order to
give you confidence that they’ve sufficiently exercised the test case in question.
You should, therefore, put them in a second stage of tests that only runs on
demand and as part of your deployment pipeline.
12.3 B i s e c t i o n
Some defects can be elusive. When I developed the restaurant system I ran into
one that took me most of a day to understand. After wasting hours following
several false leads, I finally realised that I couldn’t crack the nut only by staring
long enough at the code. I had to use a method.
12. Serialisability, here, refers to making sure that database transactions behave as though they were
serialised one after another [55]. It has nothing to do with converting objects to and from JSON or
XML.
250
12.3 Bisection
Fortunately, such a method exists. We can call it bisection for lack of a better
word. In all its simplicity, it works like this:
1. Find a way to detect or reproduce the problem.

2. Remove half of the code.
3. If the problem is still present, repeat from step 2. If the problem goes away,
restore the code you removed, and remove the other half. Again, repeat
from step 2.
4. Keep going until you’ve whittled down the code that reproduces the
problem to a size so small that you understand what’s going on.
You can use an automated test to detect the problem, or use some ad hoc way
to detect whether the problem is present or absent. The exact way you do this
doesn’t matter for the technique, but I find that an automated test is often the
easiest way to go about it, because of the repetition involved.
I often use this technique when I rubber duck by writing a question on Stack
Overflow. Good questions on Stack Overflow should come with a minimal
working example. In most cases I find that the process of producing the
minimal working example is so illuminating that I get unstuck before I have a
chance to post the question.
12.3.1 Bisection with Git

You can also use the bisection technique with Git to identify the commit that
introduced the defect. I ultimately used that with the problem I ran into.
I’d added a secure resource to the REST API to list the schedule for a
particular day. A restaurant’s maître d’ can make a GET request against that
resource to see the schedule for the day, including all reservations and who
arrives when. The schedule includes names and emails of guests, so it
shouldn’t be available without authentication and authorisation13 .
This particular resource demands that a client presents a valid JSON Web
Token (JWT). I’d developed this security feature with test-driven development
and I had enough tests to feel safe.
13. For an example of what this looks like, see subsection 15.2.5.
251
Then one day, as I was interacting with the deployed REST API, I could no
longer access this resource! I first thought that I’d supplied an invalid JWT, so
I wasted hours troubleshooting that. Dead end.
It finally dawned on me that this security feature had worked. I’d interacted
with the deployed REST API earlier and seen it work. At one time it worked,
and now it didn’t. In between these two known states a commit must have
introduced the defect. If I could identify that particular code change, I might
have a better chance of understanding the problem.
Unfortunately, there was some 130 commits between those two extremes.
Fortunately, I’d found an easy way to detect the problem, if given a commit.
This meant that I could use Git’s bisect feature to identify the exact commit
that caused the problem.
Git can run an automated bisection for you if you have an automated way to
detect the problem. Usually, you don’t. When you bisect, you’re looking for a
commit that introduced a defect that went unnoticed at the time. This means
that even if you have an automated test suite, the tests didn’t catch that bug.
For that reason, Git can also bisect your commits in an interactive session. You
start such a session with git bisect start, as shown in listing 12.9.
Listing 12.9 The start of a Git bisect session. I ran it from Bash, but you can run it in any shell
where you use Git. I’ve edited the terminal output by removing irrelevant data that Bash tends to
show, so that it fits on the page.
˜/Restaurant ((56a7092...))
$ git bisect start
˜/Restaurant ((56a7092...)|BISECTING)
This starts an interactive session, which you can tell from the Git integration
in Bash (it says BISECTING). If the current commit exhibits the defect you’re
investigating, you mark it as shown in listing 12.10
252
12.3 Bisection
Listing 12.10 Marking a commit as bad in a bisect session.

$ git bisect bad
˜/Restaurant ((56a7092...)|BISECTING)
If you don’t provide a commit ID, Git is going to assume that you meant the
current commit (in this case 56a7092).
You now tell it about a commit ID that you know is good. This is the other
extreme of the range of commits you’re investigating. Listing 12.11 shows how
that’s done.
Listing 12.11 Marking a commit as good in a bisect session. I’ve trimmed the output a little to
make it fit on the page.
$ git bisect good 58fc950
Bisecting: 75 revisions left to test after this (roughly 6 steps)
[3035c14...] Use InMemoryRestaurantDatabase in a test
˜/Restaurant ((3035c14...)|BISECTING)
Notice that Git is already telling you how many iterations to expect. You can
also see that it checked out a new commit (3035c14) for you. That’s the
half-way commit.
You now have to check whether or not the defect is present in this commit. You
can run an automated test, start the system, or any other way you’ve identified
to answer that question.
In my particular case, the half-way commit didn’t have the defect, so I told
Git, as shown in listing 12.12.
Listing 12.12 Marking the half-way commit as good in a bisect session. I’ve trimmed the output
a little to make it fit on the page.
$ git bisect good
[aa69259...] Delete Either API
˜/Restaurant ((aa69259...)|BISECTING)
253
Again, Git estimates how many more steps are left and checks out a new
commit (aa69259).
Listing 12.13 Finding the commit responsible for the defect, using a Git bisect session.
$ git bisect bad
[75f3c56...] Delete redundant Test Data Builders
˜/Restaurant ((75f3c56...)|BISECTING)
$ git bisect good
[8f93562...] Extract WillAcceptUpdate helper method
˜/Restaurant ((8f93562...)|BISECTING)
$ git bisect good
[1c6fae1...] Extract ConfigureClock helper method
˜/Restaurant ((1c6fae1...)|BISECTING)
$ git bisect good
Bisecting: 2 revisions left to test after this (roughly 1 step)
[8e1f1ce] Compact code
˜/Restaurant ((8e1f1ce...)|BISECTING)
$ git bisect good
Bisecting: 0 revisions left to test after this (roughly 1 step)
[2563131] Extract CreateTokenValidationParameters method
˜/Restaurant ((2563131...)|BISECTING)
$ git bisect bad
[fa0caeb...] Move Configure method up
˜/Restaurant ((fa0caeb...)|BISECTING)
$ git bisect good
2563131c2d06af8e48f1df2dccbf85e9fc8ddafc is the first bad commit
commit 2563131c2d06af8e48f1df2dccbf85e9fc8ddafc
Author: Mark Seemann <[email protected]>
Date: Wed Sep 16 07:15:12 2020 +0200
Extract CreateTokenValidationParameters method
Restaurant.RestApi/Startup.cs | 32 +++++++++++++++++++-------------
1 file changed, 19 insertions(+), 13 deletions(-)
˜/Restaurant ((fa0caeb...)|BISECTING)
254
12.4 Conclusion
I repeated the process for each step, marking the commit as either good or
bad, depending on whether or not my verification step passed. This is shown
in listing 12.13.
After just eight iterations, Git found the commit responsible for the defect.
Notice that the last step tells you which commit is the ‘first bad commit’.
Once I saw the contents of the commit, I immediately knew what the problem
was and could easily fix it. I’m not going to tire you with a detailed description
of the error, or how I fixed it. If you’re interested, I wrote a blog post [101]
with all the details, and you can also peruse the Git repository that
accompanies the book.
The bottom line is that bisection is a potent technique for finding and isolating
the source of an error. You can use it with or without Git.
12.4 C o n c lu s i o n
There’s a significant degree of personal experience involved in
troubleshooting. I once worked in a team where a unit test failed on one
developer’s machine, while it passed on another programmer’s laptop. The
exact same test, the same code, the same Git commit.
We could have just shrugged and found a workaround, but we all knew that
making the symptom go away without understanding the root cause tends to
be a myopic strategy. The two developers worked together for maybe half an
hour to reduce the problem to a minimal working example. Essentially, it
boiled down to string comparison.
On the machine where the test failed, a comparison of strings would consider
"aa" less than "bb", and "bb" less than "cc". That seems fine, doesn’t it?
On the machine where the test succeeded, however, "bb" was still less than
"cc", but "aa" was greater than "bb". What’s going on?
At this point, I got involved, took one look at the repro and asked both
developers what their ‘default culture’ was. In .NET, the ‘default culture’ is an
255
Ambient Context [25] that knows about culture-specific formatting rules, sort
order, and so on.
As I expected, the machine that considered "aa" greater than "bb" was
running with the Danish default culture, whereas the other machine used US
English. The Danish alphabet has three extra letters (Æ, Ø, and Å) after Z,
but the Å used to be spelled Aa in the old days, and since that spelling still
exists in proper nouns, the aa combination is considered to be equivalent to å.
Å being the last letter in the alphabet is considered greater than B.
It took me less than a minute to figure out what the problem was, because I’d
run into enough problems with Danish sort orders earlier in my career. That’s
still the shifting sands of individual experience—the art of software
engineering.
I’d never been able to identify the problem if my colleagues hadn’t first used a
methodology like bisection to reduce the problem to a simple symptom. Being
able to produce a minimal working example is a superpower in software
troubleshooting.
Notice what I haven’t discussed in this chapter: debugging.
Too many people rely exclusively on debugging for troubleshooting. While I

do occasionally use the debugger, I find the combination of the scientific
method, automated testing, and bisection more efficient. Learn and use these
more universal practices, because you can’t use debugging tools in your
production environment.
256
I n d ex
.NET, 310 A Controller, 260

1.0, 26 A/B testing, 300 impure, 273, 274
analyser, 29, 30, 67 AAA, see Arrange Act outcome, 178
Boolean default, 208 Assert activity, 13–15, 18
default culture, 255 abstraction, 69 exhausting, 190
deprecation, 220 bad, 144 involuntary, 42
ecosystem, 26 definition, 65, 100, 142, mandatory, 276
entry point, 309, 310 176, 261, 264 physical, 279
Iterator, 165 example, 69 recurring, 282
good, 146, 147 regular, 283
lightweight transaction,
249 level, 151 scheduled, 284
high, 51, 311, 312, 323 actor model, 318
package manager, 282
versus detail, 79, 320 Acyclic Dependency
SQL Server, 245
Accelerate (book), 5, 14, 37, Principle, 321
time resolution, 122
50 acyclic graph, 314
;, see semicolon Adapter, see design pattern
acceleration, 185
@Deprecated, see Accept header, see HTTP addition (arithmetic), 273
Deprecated annotation address
acceptance test, see test
[Fact], see attribute acceptance-test-driven email, see also email, 102,
[InlineData], see attribute development, 61 120, 268, 296
[Obsolete], see attribute access modifier, 319 resource, 294–296, 325
[Property], see attribute internal, 143, 319 AddSingleton method, 81
[Theory], see attribute private, 76, 206, 319 administrator rights, 299
#pragma, 67 action ADO.NET, 79, 295
80/24 rule, 134 bias, 237 affordance, 156, 157, 176
349
Index
agency, 42 public, 169–171, 196 ports and adapters, 319,

agenda reasoning about, 166, 171 323
meeting, 280 REST, 205 resilient, 299
personal, 32 benefit, 326 Tradeoff Analysis
agile process, see process client, 325 Method, xxv
AI, see artificial intelligence deployed, 252 variety, 318
Airbus A380, 17 feature flag, 209 argument, see also
airplane, 16 links, 323 parameter, 93, 242
bomber, 16 logging, 272 additional, 55
algorithm, 89, 279, 287–289 secure, 251 as context, 265
distributed system, 300 test, 324 counting, 153
inefficient, 289 user interface, 323 generated, 301
sort, 44, 45 specialised, 164 logging, 272
Ambient Context, 256 statically typed, 161, 167 mutation, 164
analyser, see tool stringly typed, 164 name, 93
Analysis Paralysis, 49 unfamiliar, 161, 185 passing, 142
annotation, 220 xUnit.net, 90 replacing constant, 88
anonymous type, 64 APL, 134 required, 156, 170, 171
antipattern, 158 application string, 57
API, 66, 156, 219 line-of-business, 4 swap, 242, 268
acronym, 156 Application Programming valid, xxvii
ADO.NET, 79 Interface, see API armed guard, 292
affordance, 156, 176 application/json, 62, 63 army
analyser, 26 apprentice, see craft air corps, 16
capability architect, xxv, 3, 5, 314 buyer, 16
advertisement, 158 architecture, 287, 288, 299, Arrange Act Assert
too many, 158 300 as scientific method, 97
change, 66, 162 component-based, 322 definition, 56
date and time, 122 conventional, 51 degenerate, 57
dependency, 219 Conway’s law, 285 purpose, 57
deprecated, 221 CQRS, 166, 299 structure, 69, 115
design, 103, 158, 160 diagram, 51 array, 304
encapsulation, 176 fractal, 151, 152, 154, 174 JSON, 205
exercise, 162 big picture, 305 params, 170
goal, 165 example, 175, 265, 312, replaced by container, 89
good, 155, 158, 171 314, 325 replacing scalar, 89
principles, 155, 168, functional core, sort, 44
176 imperative shell, 273, art
FsCheck, 304, 305 319 enjoyment of, 10
HTTP impact, 318 more than science, 65
invalid input, 93 lack of, 37, 285 of programming, 10
tools, 82 large-scale, 211 of risk assessment, 127
versus REST, 66 layered, 51, 52, 318 of software engineering,
Maybe, 146 monolith, 318 37, 220, 327
memorisation, 113 one-size-fits-all, 318 determinism, 125
object-oriented, 221 poka-yoke, 321 discomfort, 292
350
Index
example, 93, 248 Assertion Message, see code review, 28

experience, 99, 256 design pattern database, 83
shift toward Assertion Roulette, see HTTP, 82
methodology, 47 design pattern quality gate, 32
artefact, 153 asset, 47 test, 20, 82
always current, 168 assumption, 52, 248 threshold, 131
code, 181 attacker, 293–296, 298, 299 tool, 24, 25, 29, 32
inspection, 159 attention, 42, 46, 307 awareness, 42
mistake-proof, 159 to keyword, 142 of circumstances, 188
of packaging, 319 to metric, 105, 130, 132, of code rot, 154
preservation, 153 154 of quality, 132
artificial intelligence, 38 to names, 161 Azure, see Microsoft Azure
artist, 3, 10, 299 to performance, 289 Azure DevOps Services, 24,
comics, 10 to quality, 129, 154 178, 197
ASP.NET to test suite, 247
configuration, 81, 150, attribute, 59, 301, 308 B
173, 317 custom, 245 B-17, 16, 17
Decorator, 271 Fact, 90, 93 B-tree, 45
Dependency Injection, 76, InlineData, 90, 95, 103, backup, 178, 284
77, 207, 270 301, 302 backwards compatibility,
entry point, 22, 26 Obsolete, 220 219
exception, 93, 268 Property, 301 bacteria growth, 307
familiarity, 150, 310 Theory, 90, 93 balance, 56, 57, 69, 292, 296
IConfiguration interface, UseDatabase, 245 Barr, Adam, 11, 13
311 audit, 16, 267 base class, see class
Main method, 310 trail, 16, 267, 296 baseball, 33, 41, 43
Model View Controller, authentication, 251 Bash, 19, 22, 252
63, 67, 151 absence of, 37 bat-and-ball problem, 33,
options, 315 as mitigation, 296 41, 43, 44, 53, 71
web project, 21 JSON Web Token, 297 batch file, 22
Assert.True, 55 packaged, 318 batch job, 319
assertion, 65, 72, 116 two-factor, 112 BDD, see behaviour-driven
abstract, 65, 323 author, 195–198 development
append, 62, 225–227, 234 co-, 190 Beck, Kent, 4, 116, 139, 210,
collection, 72, 231, 232 code, 41 257
elegant, 72 main, 308 human mind, 115
explicit, 117 package, 282 behaviour
fail, 63 pull request, 195 addition, 203
library, 55 authorisation, 251, 311, 314, change, 162, 174, 227
message, 67 318 combined, 141
multiple, 226 automation, 19, 22, 32, 81 complex, 138
phase, 56, 69, 73, 91, 97 backup, 284 compose, 262
roulette, 62, 226 bisection, 252 correct, 248, 297
single, 226 build, 17, 18, 22, 32 desired, 119, 128, 213
superficial, 55 test, 54, 55 draining of, 81
tautological, 97, 233 checklist, 29 existing, 203
351
Index
external, 98 brain, 151, 239 outcome, 138

hidden, 204, 208, 274 capacity, 175 render, 152
incomplete, 209 compared to computer, break, 194, 238, 239,
incorrect, 268 38, 112 276–279, 285
new, 204 constraint, 45, 46, 99, compatibility, 219, 220
non-deterministic, 246, 152 contract, 66
265 emulator, 136 cycle, 321
object, 107 evaluating formal existing implementation,
of existing software, 54 statements, 41 122
proper, 73, 121 fits in your functionality, 216
test, 69, 115, 208 API, 165 breakage, 35
add, 198 application, 150 breaking change, 196,
unlawful, 181 architecture, 151, 152 219–221, 227, 282
with side effect, 266 chunk, 262, 265 brewer, 12
behaviour-driven code, 45, 46, 114, 115, bribe, 292
development, 53 135, 141, 196, 198, bridge, 12, 13, 326
behavioural code analysis, 234 Brooklyn, 117
306–308 composition, 259 Brooks, Fred, 46
benign intent, 126 essential quality, 100 browser, 314, 323
bicycle, 41, 42, 278 example, 142, 147, 175 budget, 244, 262
big O notation, 289 part, 151, 174, 318 buffer overflow, 293, 298
big picture, xxv, 305, 316 software engineering, bug, see also defect, 174,
bill by the hour, 292 46 228, 241, 247, 298
binary classification, 29 jumping to conclusions, address, 241
bisection, see also Git, 251, 43, 45, 97 despite best efforts, 201
252, 255, 256 keeping track, 15, 153, discovery, 192, 193
blog post, xxiii, xxvii, 255 265 fix, 35, 192, 204, 220, 282
boiler plate, 310 reduction, 88 struggle, 40
bomber, see airplane seven chunks, 111, 136 production, 8
Boolean side effect, 264 regression, 227, 228
default, 208 long-term memory, 111 report, 180, 246
expression, 55 misled, 44, 72 reproduction, 184, 243
flag, 144, 206, 207 motor functions, 42 tons of, 289
negation, 172 short-term memory, 39, uncaught, 252
return value, 145, 171, 133 zero, 240
261, 263 source code for, 43 build, 22, 131
value, 261 subconscious, 41, 42 automated, 17, 18, 22, 32,
bootstrap, 55, 223 tax, 264 54
Bossavit, Laurent, 192 trust, 91 configuration, 25
bottleneck, 281, 289 working memory, 131 pipeline, 246
boundary, 50, 61 branch, 81, 118, 138, 147, release, 22
HTTP, 61 152 repeatability, 272
system, 68, 69 add, 134 script, 22, 23, 32, 55, 245
test, 69, 76, 83 instruction, 39, 133 step, 22
value, 300, 301 logic, 81 build quality in, 158, 240
Boy Scout Rule, 31 on constant, 119 building a house, 3–7
352
Index
bus factor, 188, 308 syntax, xxvi Category Theory for

business sugar, 143 Programmers (book), 264
decision, 169, 299 type system, 100 CD, see Continuous Delivery
goal, xxv, 319 var keyword, xxvi cent, 41
logic, 169, 171, 246, 257, verbosity, 134 certificate, 284
322 CA1822, 139 X.509, 284
out of, 5, 36, 37 CA2007, 57 chain of command, 285
owner, 293 CA2234, 58 chair, 156, 239, 277
process, 100 cache, 267, 271 affordance, 156, 157
rule, 118, 124, 138, 290 read-through, 271 office, 157
encapsulation, 70, 72 cadastral map, 290–292 change, see also breaking
by the book, 31 calculation, 13, 33, 38, 273 change
CalendarFlag class, 207, 208 code structure, 98, 113
C call site, see also caller, 106, safe, 227
C, 293 210, 212–214 concurrent, 183
language family, xxiv, 28, caller, 142, 146, 167, 211 coupling, 306, 307
135 check return value, 147 documentation, 168, 181
C++, xxiv, 36, 44, 293 direct, 150 easy, 210
C# frequency, 307
interaction with object,
80x24 box, 135 impending, 221
99, 108, 109, 156
8, 28, 31, 92, 146, 162 in place, 215
migrate, 211, 215–219,
access modifier, 319 motivation, 53
221
analyser, 25 perspective, 7, 277
multiple, 106
branching and looping rate, 139, 257
responsibility, 108
keywords, 133 significant, 210
Campidoglio, Enrico, 19
compiler, 92, 94, 104, 144, small, 35, 61, 96
canary release, 300
220 state, 78, 164, 165
capacity
Data Transfer Object, 70 Characterisation Test, 54,
brain, 175 58
framework guidelines,
143 memory chat forum, 284
inheritance, 315 long-term, 111, 112 chatter, 285
language short-term, 136, 137, checklist, 16–18, 41
high-level, 298 141 automated, 29
verbose, 21 working, 114, 131 Command Query
like Java, xxvi of restaurant, 115–117, Separation, 166
managed code, 293 246, 249 do-confirm, 18
object-orientation, 266 hard-coded, 125, 126, engineering, 13, 37
operator 168 more than, 32
null-coalescing, 104, remaining, 123 new code, 17, 32, 54, 55,
133 of team, 192 57
null-conditional, 302 system, 293 outcome, 32, 178
overload car, 41, 42, 87 read-do, 18
return-type, 217 Carlsberg, 12 Red Green Refactor, 224
previous versions, 146 carpenter, 9 STRIDE, 292, 294
property, 143, 301, 302 case keyword, 133 surgery, 17
struct, 207 category theory, 264 take-off, 17
353
Index
team, 282 HTTP, 62, 92, 247, 325, redundant, 98, 187
warnings as errors, 25 326 removal, 237, 251
chef de cuisine, 3, 169 test, 324, 325 reuse, 40, 319
children’s book, 38 postcondition, 103 self-documenting, 161,
chunk, 115, 141, 262 SDK, 326 170
abstraction, 141, 148, cloud, 20, 21, 45, 295 shared, see code
149, 152, 175, 265 co-author, 190 ownership
code, 114, 151 coaching, xxv, 10, 88, 191 simplest possible, 52
hex flower, 312 code, see also production transformation, 88, 89,
pathways, 138, 152 code 119
short-term memory, 112, auto-generated, 25, 40, unfamiliar, 146
141 54, 72, 153 unmaintainable, xxiv
slot, 136, 148, 149 bad, 259–261 unsurprising, 310
Circuit Breaker, see design block, 139 code analysis
pattern complexity, 152 rule, 27, 28, 59, 66, 67, 75,
claim decomposition, 154, 139
role, 297, 298 155 static, 32, 57, 58, 75
class, 27 small, 134, 144, 257 driver, 53, 81
base, 122, 229, 230, 232, calling, 108, 214 false positive, 29, 60
315 dead, 7 like automated code
concrete, 213 defensive, 28, 108, 109 review, 28
declaration, 27, 268, 270, deletion, 7, 26, 27, 132, suppression, 67
310, 311 187, 238 turn on, 31
ephemeral, 19 warnings as errors, 29
delete, 221
high-level, 305 tool, 24
Domain Model, 145
high-quality, 153 code base
field, 108, 139, 153, 206
humane, 46, 174 example, see example
immutable, 265
imperfect, see also greenfield, 307
instance, 139
imperfection, 91, 123 memorise, 111
Humble Object, 123
incomplete, 204 table of contents, 314
immutable, 72, 107, 173
liability, 47 unfamiliar, 323
inheritance, 315 low-level, 154 Code Complete (book), 139,
instance, 75, 76 malicious, 126 288
member, 143, 150 metric, see metric code ownership, 187
instance, 67, 139, 142 minimal, 20 collective, 187–190, 194,
name, 163, 169, 316 multithreaded, 250 195, 197, 199
nested, 76, 232 network-facing, 293 weak, 199
private, 76 obscure, 43 code quality, see quality
sealed, 27 organisation, 43, 51, 52, code reading, 196, 197
clean code, 160 59 code review, 28, 189, 190,
Clean Code (book), 288 layer, 51 192–199, 285, 295,
client, 49, 111, 284 quality, see quality 308
API, 66 read more than written, big, 195
code, 156, 326 39, 44, 160 civilised, 197
concurrent, 183, 184 readable, 40, 41, 135, 196, initial, 193
external, 61 281 on-the-go, 190
354
Index
repeat, 197 replace with named nested, 260, 262

suggestion, 197 method, 161, 167 object, 259
code rot, see also decay, 7, stale, 161, 167, 168 pure function, 264
130, 154, 325 TODO, 67 sequential, 262–264, 266,
code smell, 25, 56, 62, 139, versus clean code, 160 274
142 Common Lisp, 36 Composition Root, see
Feature Envy, 143, 154 commons, 290 design pattern
coffee, 239 communal table, 138 comprehensibility, xxvi, 136,
machine, 239 communication, 179, 285 153
cognitive constraint, 45, 99, ad hoc, 285 computational complexity
176 arbitrary, 285 theory, 289
coherence, xxiv, 43, 186 channel, 219 computer, 7, 327
cohesion, 139, 154 face to face, 284 away from, 239, 277–279,
collapse, 265 structure, 285 285
colleague, 78, 195, 239 written, 285 compared to brain, 38,
help, 82, 198 commute, 278 39, 45, 46
collective code ownership, comparison, 123 disconnected, 292
see code ownership string, 255 in front of, 278, 279
combat aviation, 185 compartmentalisation, 114 limits, 45
combinatorial explosion, 69 compatibility, 219 personal, 11
Command, 166, 262 backwards, 219 reboot, 236
Command Query breaking, 219, 220 computer science, 44, 45,
Responsibility compiler 287, 289
Segregation, 166, 299 C#, 92, 220 education, 8
Command Query error, 25, 28, 29, 210 Vietnam, 6
Separation, 166 leaning on, 208, 210 concentration, 42
composition, 262 Roslyn, 26 concurrency, 183, 184, 191,
determinism, 264 warning, 24–30, 220 247, 249
predictability, 264 complexity, see also conference, 31
side effect, 258 cyclomatic complexity, GOTO, 158
signature, 171 46, 307 software engineering, 11
violation, 261 analysis, 153, 308 conference room, 191
command-line interface, 21 collapsed, 265 confidence, 111, 224, 227,
command-line prompt, 19 essential, 46 248, 250
command-line tool, 21 hidden, 148, 149 configuration
command-line utility, 51 increase, 129 application, 173, 174
command-line window, 19 indicator, 132 ASP.NET, 81, 173, 317
comment, 23, 180 limit, 138 feature flag, 207, 208
apology, 161 measure, 130 file, 82, 206, 208, 315
Arrange Act Assert, 56 of called methods, 141 network, 293
commit, 187 prediction, 132, 134 system, 151, 207
good, 167 structure, 151 value, 172, 173
legitimate, 167 compliance, 235 Configure method, 150, 311,
misleading, 160, 161, 196 Composite, see design 312
not all bad, 161 pattern ConfigureAwait method, 27,
pragma, 67 composition, 258, 259, 315 57–59
355
Index
ConfigureServices method, object, 87, 108 potential, 28

83, 151, 173, 312, 313 regression, 61 scare away, 296, 297
convention, 311 signing, 87 CVS, 18, 178
deleted, 27 convention, 67, 180, 311 cycle, 320–322
ConfigureWebHost method, naming, 26 life, 52
83 Conway’s law, 285 Red Green Refactor, 96,
connection string, 81–83, cooperation, 12, 284 97
151, 317 Copenhagen, 42, 278 release, 5
missing, 226 GOTO conference, 158 cyclomatic complexity, see
consciousness, 42 copilot, 18 metric
consensus, 191, 197 copy and paste, 306
lack of, 68, 73, 107 correctness, 289 D
constant, 75, 88, 89, 119 cost, 4, 21, 305 daily stand-up, 194, 275, 282
constraint, 6, 99 sunk, 195, 210, 240 format, 275
construction, 12, 13 counter seating, 117 Danish alphabet, 256
real-world, 5, 6, 13 coupling, 306, 320 Danish teachers’ union, 280
constructor, 170, 172–174, change, 306, 307 dark room, 7
207, 215–217 team, 308 data
argument, 74, 230 Covid-19, 281 access, 74, 316, 320–322
as Query, 262 CQRS, see Command component, 316
auto-generated, 72 Query Responsibility implementation, 314
overload, 170 Segregation interface, 321
parameterless, 76 CQS, see Command Query package, 321, 322
precondition, 144 Separation export, 257
side effect, 262 craft, 8–10 import, 257
validity, 106 crash, 92, 174, 268, 293, 298 persisted, 64
Constructor Injection, see airplane, 16 tampering, 293, 294
design pattern run-time, 268 version control, 305
contemplation, 26, 190, 278 creativity, 13 Data Definition Language,
contention, 183 crisps, 7 245
context, 40, 126, 160, 181 critical resource, 187 data store, 50, 61, 271
surrounding, 142, 325 cross-cutting concern, 201, data structure, 45, 70
Continuous Delivery, 5, 19, 267, 314, 315 Data Transfer Object, 69,
20, 85, 306 Decorator, 267, 271, 274 70, 315
Continuous Deployment, list of, 267 configuration, 173
204, 275, 282, 291 cross-platform, 36 role, 70
Continuous Integration, cruft, 37, 153 validation, 140, 142
131, 182, 184, 204 cruising speed, 201 versus Domain Model,
server, 20, 22, 24, 58, 182, crunch mode, 193 72, 145
246 culture, 32 data type
contract, 66 hustle, 32 built-in, 101
advertisement, 161 of quality, 154 integer, 101
design by, 99, 103 oral, 285, 290 database
encapsulation, 99 cURL, 82 access, 268, 295, 318
external, 61 customer, 36 backup, 78, 284
guarantee, 103 paying, 219 cache, 267
356
Index
cloud, 295 Debug configuration, 25 injected, 152, 260

column, 183 debugging, 256 isolation of, 68
create, 245, 246 decay, see also code rot management, 6
data structure, 45 gradual, 130, 131, 153 package, 321
design, 6 decomposition, 114, 115, polymorphic, 172
fake, 73 138 primitive, 207
graph, 78 code block, 154, 155 replace with fake, 84
implementation, 151 relative to composition, source control, 237
in-memory, 73 258, 274 stability, 283
language, 78 separation of concerns, update, 282, 283
lock, 183 274 visible, 172
logging, 272 Decorator, see design dependency analysis, 287,
permissions, 299 pattern 300, 306
query, 45, 123, 265, 289 default culture, 255 dependency cycle, 321
read and write, 268 Danish, 256 Dependency Injection
real, 73 default value, 124, 208
book, xxiii
referential transparency, defect, see also bug, 35, 88
Container, 237, 270
264 address, 241, 243, 249
configure, 270
relational, 6, 78, 151, 166, deal with later, 240
dispense with, 207
243 detect, 159
register, 76, 77, 151,
restore, 284 elusive, 250
317
row, 183 expose, 228
responsibility, 207
delete, 164 finding, 192, 251–255
Singleton lifetime, 76, 173
row version, 183 fix, 129
Dependency Injection
schema, 6, 78, 79, 257 ideal number of, 240
Principles, Practices, and
SDK, 282 in the wild, 240
Patterns (book), 207
secure, 295 introduction of, 252
platform, 298 Dependency Inversion
set up, 83
Principle, 79, 320
state, 246 prevention, 193
tampering, 294 production, 127 deployment
tear down, 83, 245 reproduction, 127, 241, automation, 19
test, 243, 245, 246 246, 250 repeatability, 272
transaction, 250 run-time, 28, 268 sign-off, 23
update, 243 troubleshooting, 235 deployment pipeline, 49, 250
DateTime struct, 72, 145, defensive code, see code establishment, 20, 23, 85
171, 211 delegation, 65, 169, 174, 175 issue, 20
daughter, 280 DELETE, see HTTP Deprecated annotation, 220
DDD, see domain-driven delimiter, 135 deprecation, 220, 221
design demonstration flight, 16 describing a program, 6
DDL, see Data Definition denial of service, 293 design
Language distributed, 293, 298 by contract, 99, 100, 103,
DDoS, see denial of service dependency 109
dead code, 7 change frequency, 283 error, 158
deadline, 192, 193 composition, 207 design pattern, 224, 279, 300
deadlock, 58 external, 243 Adapter, 315
Death Star commit, 186 formal, 172 Assertion Message, 67
357
Index
Assertion Roulette, 62, Devil’s Advocate, 119–121, evolution, 100

226 123–125, 128 motivation, 145, 169
Circuit Breaker, 267, 271 DI, see Dependency domain name, 284
Composite, 259 Injection domain-driven design, 53,
Composition Root, 322 diff, 179, 181 169
Constructor Injection, 75, tool, 238 domain-specific language,
172, 310 Dijkstra, Edsger, 11 78
Decorator, 267–271, 274, diminishing returns, 191 Don’t repeat yourself, see
317 directed graph, 227 DRY principle
Humble Object, 80, 123, directory, see also done done, 192
242 subdirectory, 19, 314, 315 door handle, 156
Iterator, 165 discipline dot-driven development, 158
Model View Controller, academic, 288, 305 dotnet
63, 67, 151, 311 engineering, 10, 11, 13, build, 22, 55
Null Object, 76 14, 31 test, 55
Repository, 74 future, 327 double-blind trial, 237
Unit of Work, 249 esoteric, 34 double-entry bookkeeping,
Value Object, 72 intellectually demanding, 53, 91, 224
Visitor, 160 33 driver
Design Patterns (book), 259 discomfort, 292 code analysis, 75, 81
design phase, 6 discoverability, 158 code as answer to, 88
desktop application, 293 discriminated union, see example, 53
Deursen, Steven van, 207 sum type extrinsic, 53
developer, see also software discussion multiple, 76
developer repeated, 285 of behaviour, 68
as resource, 5 technical, 284 of change, 53, 103
back-end, xxiv written, 284 of implementation, 61
in a hurry, 319 disillusionment, 10 of transformation, 91
main, 199, 289 disjoint set, 139 test, 74, 115, 224
original, 284, 289 do keyword, 133 integration, 208
remote, 194 document database, see driving, 41, 42
responsibility, 295 database Dronning Alexandrine’s
single, 192 documentation, 59, 60, 160, bridge, 12
Visual Studio, 30 167 DRY principle, 107
development, see also high-level, 168 DSL, see domain-specific
software development online, 28 language
back-end, 9, 188 rule, 57, 58 DTO, see Data Transfer
front-end, 9 scalability, 280 Object
greenfield, 2, 203 stale, 41, 168, 196 duplication
individual, 231 doing dishes, 278 address, 147, 234
user-interface, 188 dollar, 33 look out for, 52
development environment, domain, 148 needless, 197
see also IDE, 135, 157, Domain Model, 70–72, 74, of implementation code,
168 315, 318–322 91
development machine, 22, abstraction, 79 test code, 61
82, 205, 255, 256 clean, 79 validation, 144
358
Index
DVCS, see version control Data Transfer Object, 70 equilibrium

system good, 156 unstable, 153
dyslexia, 181 invariants, 144 error
misunderstood, 108 compile-time, 92, 160
E poor, 173, 223 finding, 255
economics, 8, 132 purpose, 165, 169 pilot, 16
edge state, 106 programmer, 243
of system, 265, 266 versus strings, 58 report, 93
edge case, 69 enclosure diagram, 307, 308 reproduction, 243
editor, see also IDE, 13 engineer, see also software spelling, 25
vertical line, 135 engineer, 3, 12, 13, 177 error message, 17, 18
education chemical, 12 essay, 160
computer science, 8, 44 real, 13, 177, 199 essence, 141, 146
self, 280 engineering, see also ethics, 292
effort, 16 software engineering, 29, eureka, 278
continual, 35 33 exception, 301
heoric, 210 deterministic process, 327 ArgumentNullException,
little, 42 discipline, 10, 11, 13, 14, 92
mental, 42, 43 31 ArgumentOutOfRange-
small, 32 future, 327 Exception, 300,
Eiffel (language), 100 mechanical, 44 301
elevated privileges, 236 handling, 92
method, 13, 17
elevation of privilege, 293, message, 93, 107, 180
practice, 199, 210
299 NotImplementedExcep-
real, 326
email tion, 213
relationship with science,
address NullReferenceException,
44, 98
as identification, 102 92
security, 300, 308
bogus, 102 run-time, 92
English, 181
validation, 102 versus compiler error,
US, 256
confirmation, 229 160
Entity Framework, 79 type, 92
grammar, 180
personally identifiable entry point, 52, 150, 309, 310 unhandled, 93, 127, 268
information, 294, 296, ASP.NET, 22, 26 exclamation mark, 91, 92,
297 environment 94, 96
procrastination, 244, 277 concern, 326 execution
unit test, 229, 230, 303 configuration, 82 branch, 119
employee, 297 data, 264 path, 88, 115
hire, 282 development, 135, 157, repeatability, 272
new, 113 168 exercise, 284, 316
regular, 194 pre-production, 20 API design, 162, 163
empty string, 57, 103, 105 production, 85, 127 physical, 278, 279
emulator, 39, 136 debugging, 256 experience, 37, 125
encapsulation deployment, 23 accumulated, 9, 10
broken, 144 lack of, 20 individual, 11, 13, 93, 99,
business rule, 70, 72 programming, 24, 25 235, 256
contract, 87, 99 Equals method, 72 personal, 255
359
Index
professional, xxiv, 44 fee, 28 Fowler, Martin

subjective, 42 reservation, 296 code that humans can
experiment, 15, 97, 236, 237, feedback, 49, 52, 60, 160 understand, 45, 176
241 feudalism, 290 Data Transfer Object, 70
Git, 18, 185, 186 Fiddler, 82 quality, 35, 37, 40, 47
result, 11 file, 314 Strangler, 211
code, 315 FP, see functional
F dirty, 231 programming
F#, 134, 146, 160, 322, 323 executable, 318 fractal architecture, see
Fact attribute, 90, 93 organisation, 314 architecture
fail fast, 103 filter, 122, 123, 262, 314, 315 fractal tree, 151, 152
failure, 148, 210, 247 finite state machine, 300 fractals, 151, 154
single point of, 187 firefighting, 193, 275 framework
Fake Object, see Test Double Firefox, 314 automated testing, 14
FakeDatabase class, 69, 73, first language, 181 data-access, 52
74, 83, 122, 212, 213 fits in your head, 150–152, experience with, 311
fallacy 154, 176, 262, 274, 312 familiarity, 309, 310
logical, 37 API, 165 MVC, 63, 67
sunk cost, 195, 210, 240 architecture, 314 not-invented-here
false negative, 97, 226, 247, chunk, 262, 265, 312 syndrome, 6, 36
248 code, 114, 115, 135, 198, security, 271
false positive, 29, 60, 247 309 unit testing, 55, 90, 305
falsifiability, 97, 237 composition, 259 Freakonomics (book), 132
fault tolerance, 267, 271, 300 criterion, 196 Freeman, Steve, 7
feature, 50, 52, 192, 193 evaluation, 141, 311, 317, frequency, 283
add, 35, 40, 129 323 change, 307
big, 220 example, 142, 147, 175 frog
completion, 193, 208 object, 100 boil, 130
configuration, 208 part, 174, 318 fruit
cutting across, 267 system, 114 low-hanging, 24
delivery, 276 flag, see also feature flag, FsCheck, 301–305
deployed, 201 144, 206, 208 NegativeInt, 302
difficult, 204 flow, see also zone, 42, 277 NonNegativeInt, 302, 304
done, 192 focus, 16, 244 PositiveInt, 302
end-to-end, 52 Foote, Brian, 153 function
incomplete, 204 for keyword, 133 self-contained, 46
new, 204, 209, 220, 282 foreach keyword, 133 functional core, imperative
optional, 28 forensics, 40 shell, 266, 273, 318
security, 251, 252, 271 forgetfulness, 16, 38 example, 319
subdirectory per, 314 formatting, 187, 196, 198 in presence of
suggestion, 278 blank line, 56 object-oriented code,
feature branch, 220 culture, 256 274
Feature Envy, see code smell Git commit, 179, 180 functional programming, 14,
feature flag, 184, 204, guard, 24 238, 264, 266
206–209, 220 line width, 136 influence on C#, xxvi
configuration, 208 foundation, 6, 11, 19 FxCop, 26
360
Index
G graphical user interface, greater than, 123, 255, 256

Gabriel, Richard P., 36 18, 19, 180 greater than or equal, 123,
game programming, 9 HEAD, 231 302
Gantt chart, 6 history, 59, 186, 187 grocery store, 279
gardening, 3, 7, 8 init, 19 ground level, 34
Gawande, Atul, 16, 17 issues, 18 Growing Object-Oriented
generics, 146, 215 learning, 18 Software, Guided by Tests
nested, 216 log, 179, 283 (book), 7, 61, 78, 325
geographical survey, 127, 128 master, 184, 185, 187, 197 growing season, 290
geometry, 127 deployment, 23 guarantee, 87, 103, 108
GET, see HTTP incomplete feature, 204 guard
GetHashCode method, 72 merge, 197, 198, 213, 216, armed, 292
getter, 108, 143 218 Guard Clause, 100, 139, 145,
GetUninitializedObject online service, 19, 197 187, 262
method, 107 push, 186 natural numbers, 101, 102
Gibson, William, 14, 327 reason for using, 182 null, 75, 94
Git, 14 rebase, 1 GUI, see graphical user
.git directory, 19 repository, 19, 255 interface
50/72 rule, 179, 180 local, 184 GUID, 264, 265, 294
Bash, 19, 252 secrets, 82 guidance, 11, 21
basics, 18 stage, 233 guideline, 10, 26, 88, 89, 103
bisect, 251–255 stash, 186, 231, 232, 241 guild, see craft
blame, 40 tactics, 178 guitar, 10
branch, 184–187, 197 user-friendliness, 18
command line, 19, 179, Git flow, 197 H
180 GitHub, 19, 178, 197, 284 hack, 32, 92, 131, 321
command-line interface, GitHub flow, 197, 198 hammer, 291
18 GitLab, 178 happy path, 64
commit glucose, 43 hard drive, 19, 112, 186
big, 186 Go To Definition, 315, 317 hard limit, 138
empty, 19 Go To Implementation, 317 hard-coded
five minutes from, 218 God Class, 158 capacity, 125, 126, 168
hidden, 231 Goldilogs, 272 constant, 89
ID, 253 GOOS, see Growing path, 66
self-explanatory, 180, Object-Oriented return value, 61
181 Software, Guided by value, 75, 77, 78, 83, 87,
small, 185 Tests (book) 88, 303
commit message, 167, GOTO conference, 158 hardware, 14, 21, 290
168, 178–182, 198 government, 35 control of, 126
co-author, 190 grammar, 180 hash, 183
empty, 178 graph hash index, 45
connection string, 82 acyclic, 314 Haskell, 160, 274
database schema, 79 directed, 227 absence of null references,
de-facto standard, 18, 178 graph database, 78 146
experimentation, 185 graphical user interface, 18, big function, 134
game changer, 231 19, 82, 164, 180 category theory, 264
361
Index
learning, 266 house, 3, 5, 87 I

linter, 25 House, Cory, 196 IConfiguration interface, 82,
Maybe, 146 HTTP, 318, 322, 326 311
QuickCheck, 301 200 OK, 55 IDE, 14
side effect, 166 201 Created, 55 acronym, 281
HDMI, 159 204 No Content, 115 file view, 316
head waiter, see maître 400 Bad Request, 93 guidance, 21, 170, 281
d’hôtel 403 Forbidden, 297, 298 navigation, 310, 315
headline, 180 500 Internal Server Error, refactoring, 223, 227,
height restriction barrier, 93, 94, 116, 226 234
159 boundary, 61 use to compile, 22
hello world, 54, 55, 61 client, 247 if keyword, 133
helper method code, 319 illegal states
extract, 139, 234 content negotiation, 62 unrepresentable, 159
motivation, 66 DELETE, 294, 296 illusion
Henney, Kevlin, 6 GET, 205, 251, 293, 296, maintainability, 26
heroism, 210 297, 323 immutability
heuristic, 10 header, 325 class, 72, 107, 173
API design, 155 field, 265
Accept, 62
Arrange Act Assert, 56, object, 106
Content-Type, 62, 63
69, 115 imperative mood, 18, 179,
Location, 295
for first feature, 64 180
interaction, 205, 209
hex flower, 137, 138, 141, imperfection, 91, 106, 123,
POST, 78, 82, 294, 323,
142, 148–151, 312, 313, 126
325
317 implementation detail, 172,
PUT, 242, 294
hexagon, 137, 138, 142, 150 264
Hickey, Rich, 46, 238 request, 21, 67, 151, 312
coupling, 107, 320
hierarchy, 314 logging, 272
Dependency Inversion
directory, 315 response, 226, 297 Principle, 79
inheritance, 315 content, 226 hidden, 165, 170
of communication, 167, logging, 272 irrelevant, 176
179, 219 specification, 116 unknown, 99, 171, 174
rigid, 285 status code, 55, 65, 94, view, 316
type, 227 116, 261 improvement
hill, 34 error, 115 heuristic, 119
hipster, 117, 138, 168 verb, 66 loss of ability, 35
history, 12 HttpClient class, 324, 325 impure action, 273, 274
line width, 135 HTTPS, 295, 296 incantation, 236
of software development, humane bounds, 154 incentive
14, 15 humane code, 46, 174 perverse, 132, 292
rewrite, 1, 19 Humble Object, see design indentation, xxv, 271
HIV, 29 pattern infinity, 152
Hoare, Tony, 11 hunt-and-peck typing, 281 information disclosure, 293,
HomeController class, 63, hypermedia controls, 66, 205 296
207 hypothesis, 37, 97, 236, 237, infrastructure
hotspot, 307, 308 241 cloud, 45
362
Index
code, 204 interpersonal, 197 managed code, 293

digital, 35 object, 99, 103, 108 null, 146
inheritance, 315 social, 177 JavaScript, 25, 282, 298
single, 315 Interactive Development Jenkins, 24
initialisation, 106 Environment, 281 job security, 113
object, 144 interception, 269, 295 journeyman, 9, 10
InlineData attribute, see interface JSON, 250
attribute add member, 122, 213 array, 205
inlining, 290 affordance, 156 configuration, 315
input, 50, 103 cycle, 320 document, 61, 70, 83, 92,
acceptable, 103 delete member, 214 173
invalid, 93, 95, 100, 103 extra method, 212 object, 64
logging, 273 go to implementation, parsing, 326
malevolent, 293 315 representation, 205, 323,
null, 99 versus base class, 229 325
parsing, 147, 148 internal, see access modifier response, 61, 62
query, 50 Internet, 14 serialisation, 64, 325
required, 156 internet JSON Web Token, 251, 252,
valid, 100 disconnected from, 292 282, 297, 298
validation, 77, 92, 115 interpreter, 180 redaction, 272
insight, 210, 239, 268, 278 introvert, 190 judgment, 37, 220
inspiration, 17, 44, 279 intuition, 33, 38, 41 human, 326
instruction, 17, 160 invariant, 80, 109, 144, 145, moral, 31, 32
instrumentation, 267, 272 156 subjective, 53, 79
insurance, 292 investigation, 16, 241, 307 jumping to conclusions, 43,
intangible, 13, 44, 291, 292 IPostOffice interface, 230, 45, 97
integer, see also number, 100 231, 233, 271, 317 JWT, see JSON Web
16-bit, 101 IReservationsRepository Token
8-bit, 101 interface, 73, 74, 76, 77,
default value, 124 79, 81, 83, 121, 123, 156, K
non-negative, 302 161–163, 211–214, 269, Kahneman, Daniel, 42, 43
non-positive, 301, 302 271, 316, 317 Kanban board, 275
signed, 102 IRestaurantManager kata, 280
unsigned, 102 interface, 260, 261 Kay, Alan, 11, 12
Integrated Development IT professional, 293, 295, keyboard, 191, 281
Environment, see IDE 298 keyboard shortcut, 315, 316
integration test, see test Iterator, see design pattern king, 290, 291
IntelliSense, 157 King, Alexis, 147
intent, 163, 167, 196, 198 J KISS, 238
benign, 126 Java kitchen, 117, 118, 168, 169
code, 161 deprecation, 220 kitchen timer, 277
interaction developer, xxiv knowledge
external world, 145, 229 example code, xxiv, xxvi existing, xxiii, xxiv, 11
hidden, 262 high-level language, 298 expansion, 280
HTTP, 205, 209 inheritance, 315 local, 290
IDE, 281 like C#, xxvi loss of, 285
363
Index
packaged, 26, 45 legibility, 290–292 LoggingReservations

painstakingly acquired, less than, 123, 255 Repository class, 270, 271
114 less than or equal, 123 logistics, 5, 13
knowledge distribution, 308 liability, 87 long hours, 193, 279
knowledge gap, 288 code, 47 loop, 133
knowledge map, 308 library tight, 289
knowledge silo, 190, 194 JSON Web Token, 282 lottery factor, 188
knowledge transfer, 191 mock object, 237 low-hanging fruit, 24
Knuth, Donald, 11 open-source, 219 LSM-tree, 45
reusable, 45, 58 Lucid, 36, 40
L life cycle, 52 lunch, 194, 282
lab coat, 237 light, 7, 9, 210
lambda expression, 270 line M
land, 87, 290 blank, 56 maître d’hôtel, 102, 169, 246
ownership, 290 Arrange Act Assert, authentication, 297
language, see also 56, 57, 69, 115 schedule, 251, 293, 296,
programming language Git commit message, 304
familiarity, 309 179 machine code, 290
first, 181 section, 139 machine learning, 9
latency, 190, 192 vertical, 135 magic spell, 236
later is never, 240 wide, 271 Main method, 265, 309, 310
LaTeX, xxviii, 4 line break, 23, 271 maintainability
law of unintended line width, xxvii, 135 illusion, 26
consequences, 132 line-of-business application, maintainer, 189
layer, 52, 320 4 maintenance burden, 214
layered architecture, see lines of code, see metric maintenance mode, 4
architecture LINQ, 122, 124, 125 maintenance task, 221
leader linter, 24, 25, 30, 32 maintenance tax, 212
technical, 132 as driver, 53, 76 MaitreD class, 169–174, 304
leadership, 300 false positive, 29 man-in-the-middle attack,
lean manufacturing, 159 warnings as errors, 29 293, 295, 296
lean on the compiler, 210 Liskov Substitution management, 191, 194
lean software development, Principle, 227 manager, 31, 177, 178, 210,
158, 240 listen to your tests, 325 292
Lean Startup (book), 50 literal, 59 non-technical, 31, 292
left to right, 122 literary analysis, 160 manoeuvrability, 185, 231,
legacy code, 111, 223 localhost, 205 233
avoid, 114 Location header, see HTTP manufacturing, 327
deliberate, 129 locking lean, 159
escape, 114 optimistic, 183, 184 Martin, Robert C.
gradual decay, 153 pessimistic, 183 abstraction, 65, 100, 142,
memory, 113, 136 log, 93, 267, 268 176, 261, 264
programmer, 113 log entry, 174, 270, 271 Transformation Priority
realisation, 129 logging, 77, 267, 268, Premise, 88, 89
refactoring, 114 270–273, 315, 317, 318 triangulation, 119, 123
legacy system, 211 LoggingPostOffice class, 271 mason, 5
364
Index
master, see craft method milliseconds, 289

materialised view, 299 signature, 164 mindset
mathematics, 33, 41, 264 method call engineering, 14
fractals, 151, 152 blocking, 102 team, 132
matryoshka dolls, 268, 269 methodology, xxv, 15, 256 tinkering, 36
Maybe, 146, 162 deliberate, 56 minimal working example,
measure, 37, 97, 246, 290 engineering, 13 251, 255, 256
triangulation, 127 lack of, 10 mistake
measurement, 291, 292 quantitative, 327 all the time, 41
performance, 267 scientific, 97
cheap, 185
proxy, 292 software development,
commit, 185
triangulation, 127, 128 47, 53
easy to make, 53, 224
medieval village, 290 software engineering, 50
hide, 186, 201
meeting, 235, 275, 280 metric
attention, 130, 154 prevention, 126
memorisation, 111–114 proof, 158
memory cyclomatic complexity,
130–133, 147, 149, 150, reduce risk of, 88
aid, 17 repeat, 243
306
fading, 112
example, 136, 138–140, typing, 281
long-term, 111–113, 136,
174, 260, 262, 311, misuse, 158
141
312, 317, 323 mitigation, 292, 295, 297,
short-term, 111–113, 154
explicitly consider, 152 298
capacity, 136, 137, 141
of called methods, 140 mob programming, 184,
chunk, 141
one, 242 191, 192, 199, 316
hexagonal layout, 142
seven, 46, 105, 169 driver, 316
limit, 39, 46, 99, 133
threshold, 136 mobile phone app, 293
magical number seven,
Visual Studio, 105 mock, 73, 107
39
depth of inheritance, 105 Model View Controller, see
slot, 136, 138, 149 invent, 132 design pattern
unreliable, 38, 39 lines of code, 132, 134 module, 189, 258, 314
working, 39, 111, 114, 131 attention, 154
memory footprint, 289 money, 12, 21, 35
example, 139, 140, 174, monolith, 219, 318, 319, 323
merge conflict, 183 260, 312, 317, 323
merge hell, 182–184, 220 morals, 31, 32
explicitly consider, 152
merge sort, 45 morning, 194, 275, 280
Visual Studio, 105
metaphor, 3–8, 38, 97 motivation, 53, 57, 309
monitor, 130
accountant, 8 Domain Model, 145
practicality of, 132
author, 8 useful, 132 extrinsic, 53
brain, 38, 112 Visual Studio, 133 intrinsic, 38
gardening, 7, 8 Meyer, Bertrand, 100, 166 package, 321
house, 4–8 micro-commit, 187 process, 178
Russian matryoshka micro-service, 318 rule, 28
dolls, 269 microseconds, 289 motor function, 42
software craftsmanship, 9 Microsoft, 28, 72, 78, 293 multi-tenancy, 269, 323, 325
triangulation, 127 Microsoft Azure, 268 mutation, 106
metering, 267 Milewski, Bartosz, 264 artefact, 153
365
Index
MVC, see Model View null-forgiving operator, object-oriented language,

Controller 144 xxiv, 146, 266
myopia, 37, 192, 255 nullable reference type, object-oriented
28, 72, 92, 146, 162 programming, 14, 100,
N alternatives to, 146 108, 211, 238
naming convention, 26 gradually enabling, 31 object-relational mapper, 78,
nanosecond, 122, 289 suppression, 144 79, 243, 320, 321
NASA, 11 NullReferenceException, reinvention, 6
NATO, 11 92 versus SQL, 238
natural number, see number return value, 161 obligation, 87, 108
navigation, 24, 111, 151, Null Object, see design Obsolete attribute, see
180, 314 pattern attribute
need it later, 215 NullRepository class, 76, 77, Occurrence class, 215–218
negative number, see number 81 office, 279
nested class, see class number, see also integer home, 284
nesting, 259, 260, 262, 274 128-bit, 294 open, 284, 285
dolls, 268, 269 increment, 133 own, 284
object, 268 natural, 100–102, 106, 300 one-time code, 112
nihilism, 10 negative, 102, 107, 300, open-source software, 278,
nil, 89 301 285
no-op, 67 one, 133 OpenAPI, 205
Nobel laureate, 42 positive, 102, 108
opening hours, 168, 169
noble, 290, 291 random, 273
operations specialist, 177
non-breaking change, 219 seven, 39, 46, 111, 131,
operations team, 78
non-determinism, 265, 266, 133, 138
operator
273 ten, 41, 43
greater-than, 123
non-nullable reference type, zero, 107
greater-than-or-equal,
see null number-line order, 122
123
Norman, Donald A., 156, NUnit, 301
less-than, 123
157
O less-than-or-equal, 123
NoSQL, 78
object minus, 302
notification area, 277
NPM, 282 composition, 259, 315 null-coalescing, 104, 133
NuGet, 282 equality, 72 null-forgiving, 92, 144
null, 28 immutable, 106 ternary, 104
ArgumentNullException, polymorphic, 268 unary, 302
92 shared, 76 Option, 146
check, 92, 99 object-oriented API, 221 order
coalescing operator, 104, object-oriented code, 238, ascending, 122
133 266, 274 ordering, 227
Guard Clause, 75, 81, 94, object-oriented organisation
96, 98 composition, 258, 259 healthy, 32
nil, 89 object-oriented rhythm, 193
non-nullable reference decomposition, 274 unhealthy, 32
type, 28, 99, 106, 144, object-oriented design, 139, ORM, see object-relational
162 142, 160, 259, 274 mapper
366
Index
outcome how many, 153 PII, see personally

actual, 56, 97 query, 50 identifiable information
adverse, 127 swap, 242, 243, 268 pilot, 16–18
direct, 178 Parameter Object, 153 test, 16
expected, 56, 72, 97, 116 parameterless constructor, pipeline, see deployment
falsifiable, 97 see constructor pipeline
improvement, 17, 29, 32 Parametrised Test, see test pixel, 258
negative, 178 params keyword, 170 plain text, 54
positive, 178 parsing, 144, 145, 147, 148, document, 61
predicted, 97 173 planning, 5, 6, 13, 49
quantitative, 97 partial function, 148 platform, 258, 309
successful, 13, 124 password, 296 defect, 298
versus process, 178 pattern language, xxiii plot of land, 87, 290
output, 70, 100, 103 pause point, 16 poka-yoke, 159, 321
indirect, 229 peasant, 290 active, 159
parsed, 148 performance, 201, 287–290, passive, 159
terminal, 252, 253 292 policy, 21, 198
to input, 262, 264 fixation, 292 politeness, 198
type, 163 issue, 58 polymorphism, 146, 172,
over-engineering, 215 performance monitoring, 229, 268
overbooking, 246, 247, 268 267 Pomodoro technique, 276,
test, 116, 117, 226 permission, 198, 293, 299 277
overload, 213 persistent storage, 77 pop culture, 12
add, 212, 213 personal computer, 11 ports and adapters, 318,
return-type, 217 personally identifiable 319, 323
overlogging, 272 information, 296 positive number, see number
overtime, 192 perverse incentive, 132, 292 POST, see HTTP
petri dish, 307 PostAsync method, 66
P phase, 5 postcondition, 226–228
package, 30, 283, 318, 319, act, 56, 57, 69, 97, 115, contract, 108
321–323 124 guarantee, 108
author, 282 arrange, 56, 69, 115, 124 invariant, 109, 144
data access, 321, 322 assert, 56, 73, 91, 97 Postel’s law, 103
distribution, 282 construction, 5, 6 weaken, 228
encapsulation, 156 design, 6 Postel’s law, 103, 106, 109
reusable, 44, 45, 301 green, 97, 98, 104, 107, Postel, Jon, 103
test, 322 125 Postman, 82
update, 282, 283 programming, 5 PowerShell, 22
version, 282 red, 97, 107, 125 precondition, 143, 145, 156
package manager, 282 refactor, 96, 98, 104 check, 105, 144, 145
package restore, 237 phone number, 112, 113 contract, 108
pair programming, 189–192, physical activity, 279 invariant, 109, 144
199, 295 physical design, 4 Postel’s law, 103
rotation, 190 physical object, 12, 13, 156 responsibility, 108
parameter, see also physical work, 279 strengthen, 232
argument, 152 physics, 44 weaken, 212, 227
367
Index
predicate, 260, 262, 263 measure, 279 getter, 143

predictability, 264 metric, 132 read-only, 72, 75, 172, 176
prediction, 37, 97, 236, 237 negative, 279 Visual Basic, 301
PRINCE2, 276 personal, 279, 285 Property attribute, see
private, see access modifier tip, 280 attribute
probability, 127 profit, 35 property-based testing, see
problem Program class, 22, 26, 27, test
address, 236, 237 150 prophylaxis, 134
alternative solution, 9 programmer, see also prose, 180, 181
dealing with, 236 developer Pryce, Nat, 7
detect, 252 good, 45, 176 psychology, 42
disappear, 236 irreplacable, 113 pull request, 197, 198, 285
explaining, 239 legacy code, 113 big, 134, 194, 198
manifestation, 236, 272, maintenance, 105, 309 punch card, 289
278 other, 177, 310 punctuation, 180
reaction, 236 responsibility, 295 pure function, see also
reproduction, 246, 251 single, 192 referential transparency,
solving, 235, 238 suffering of, 129 237, 264–266, 273, 274
stuck, 238 third-party, 326 PureScript, 166
unanticipated, 193 user-interface, 188 purpose, 36, 37, 44, 258, 296
process, 275, 291 programming by PUT, see HTTP
agile, 284 coincidence, 236, 237 puzzle, 33, 43
approval, 190 programming language
compilation, 167 advanced, 14 Q
external, 299 C-based, 135 quality, 129, 301
formal, 308 components, 258 build in, 159, 240
iterative, 197 cross-platform, 36 essential, 100
long-running, 102 density, 134 internal, 31, 35, 37, 98,
mistake-proof, 159 emulator, 39 154
subconscious, 279 functional, 266 better, 131
versus outcome, 178 high-level, 298 low, 40
procrastination, 276 keyword, 133 quality gate, 31, 32
product owner, 177 layout, 135 quantifiable result, 36, 97
production code learning, 18, 279, 280 Query, see also Command
as answer to driver, 88 mainstream, 320, 321 Query Separation, 166
bug, 228 new, 40 composition, 262
change, 224, 228 statically typed, 157, 161 constructor, 262
confidence, 224 tools, 24, 25 deterministic, 264, 265
coupled to test code, 228 verbosity, 21, 134 example, 176, 262, 263
edit, 203 progress, 12, 14, 35, 45 favour, 166
refactoring, 227, 229 project, 4–6 non-deterministic, 264,
rule, 58 project management, 6, 37 266
productivity, 191, 235, 278, proper noun, 256 parameter, 50
281, 285 property, 301 side effect, 261, 262
deleting code, 132 C#, 301, 302 type, 163, 171
long hours, 279 declaration, 72 queue, 50, 249, 299
368
Index
QuickCheck, 301 Rename Method, 218, 227 risk, 299

quicksort, 45 Rename Variable, 227 risk assessment, 127
safe, 227, 228 robot, 156
R test, 231 industrial, 258, 327
race condition, 246–248 test code, 224, 232, 234 role
Rainsberger, J.B., 47 apart, 229 claim, 297, 298
RAM, 39, 45, 112 to property-based test, object, 70
random number, 273 301–303 rollback, 246
random number generator, upon rot, 325 roof, 6
264 toward deeper insight, roofer, 9
random value, 301, 302 209 room
range, 148, 212, 213 Refactoring (book), 143, dark, 7
readability, 41, 281 223, 224, 227 root cause, 255
code review criterion, 196 reference type, see also null, Roslyn, 26, 29
nudge, 135 28 rotation, 56, 57
optimise for, 40, 79 referential transparency, 264, routine, 194, 279
reader 265, 273 routing, 151, 311
future, 59, 160, 163 regression, 227, 228 rubber duck, 239, 240, 251
readme, 168 likelihood, 126 rubber stamp, 194, 198
real world, 29, 100, 258 prevention, 55, 61, 127, Ruby, 89, 282
reality, 6, 10, 52, 192, 290 243 RubyGems, 282
physical, 6 relationship type, 206 rule
reboot, 236, 237 release, 219–221 against decay, 131
receiver, 160 canary, 300 analyser, 26, 27, 30, 31,
recursion, 89 Release configuration, 22, 25 57, 139
Red Green Refactor, 96, 97, release cycle, 5 breaking, 131
125, 128, 224 repeatability, 272 business, 118, 124, 138,
execution time, 244 repetition, 251 290
red phase, 103, 107 Repository, see design encapsulation, 70, 72
Reeves, Jack, 5, 13 pattern Command Query
refactoring, 98, 203 repudiation, 293, 296 Separation, 166
Add Parameter, 228 research, 5, 38 disable, 60
backbone of, 224 resiliency, 298–300 documentation, 58
big, 220 REST, 66 extra, 264
candidate, 139 restart, 236 formatting, 256
code ownership, 187 restaurant owner, 294, 296 hard, 132
commit, 229 RESTful, 205 line height, 135
database, 245 RESTful Web Services machine-enforced, 31
Extract Method, 187, Cookbook (book), 116 motivation, 28
227, 228 return on investment, 299 redundant, 132
IDE, 234 revelation, 278 threshold, 131, 132
Inline Method, 187 review, 13, see code review versus food for though, 89
legacy code, 114 reviewer, 195–198 rule of thumb, 10, 182, 204,
Move Method, 143, 227 rework, 220 210, 242
opportunity, 125 Richardson Maturity running, 278
prophylactical, 134 Model, 66 Russian dolls, 268, 269
369
Index
S serialisation, 64, 249, 250, literary composition, 160

sabotage, 119, 233 325 situational, 8
safety net, 224, 228, 234, 244 server, 21 specialised, 9
salary, 21 setter, 108, 143 troubleshooting, 235
scaffold, 20 seven, 46, 136–138, 151–154 slice
scalar, 88, 89, 119 magical number, 39 vertical, 49–52, 54, 60, 61,
schedule proxy, 46 77
certificate update, 284 threshold, 130, 131, 133 first, 64, 85
package update, 283 token, 39, 133 happy path, 64
synchronisation, 190 shared code, see code purpose, 64
team, 282 ownership small step, 61, 88, 194, 220
shell script, 22 SMTP, 102
school, 160, 177, 280, 281
shifting sands of individual snapshot, 183, 184, 187
science, 44, 97, 98
experience, 11, 13, 93, 99, social media, 258
scientific evidence, 13
235, 256 software
scientific method, 97, 236,
shopping, 279 reusable, 45
237, 256
shower, 278 successful, 4
scientist, 3, 44
side effect, 162, 164–166,
screen, 42, 239, 258, 281 sustainable, 67
171, 258, 259, 261–266
Scrum, 276, 283 unsuccessful, 4
constructor, 262
sprint, 283 software craftsmanship,
Haskell, 323
retrospective, 282 8–10
hidden, 43
SDK, 282, 326 software crisis, 11, 14
logging, 273
sealed keyword, 27 software developer, see also
sign-off, 13, 23, 198, 199
seating developer
signal, 29, 247
bar-style, 118, 168 collaboration, 189
signature
counter, 117 professional, 31
digital, 296
overlap detection, 174 skill, 8
method, 145, 146,
second, 138, 168, 169 software development
162–164, 166, 170,
single, 117, 138, 168 171 asynchronous, 285
security, 271, 287, 288, 290, identical, 213 highest-ranked problem,
292, 300 Simple Made Easy 182
balance, 296 (conference talk), 238 history, 14, 15
mitigation, 292 simplest thing that could industry, 9, 13, 14, 45
security by obscurity, 294 possibly work, 75, 117, age, 3
Seeing Like a State (book), 215 improvement, 8
290 simplicity, 46, 238 management, 292
self-hosting, 55, 324 simulation, 13, 21 process, 52, 276
self-similarity, 154 single point of failure, 187 latency, 192
Semantic Versioning, 218, SingleOrDefault method, regular, 35
219 124–127 professional, 29, 235
semicolon, 135 Singleton lifetime, see reality, 203
sender, 160 Dependency Injection project
sensitivity, 231, 290 skill, 199 bad, xxiv
separation of concerns, 257, decomposition, 155 sustainable, 40
268, 274, 314 legacy, 113 team, 177, 287
370
Index
software engineering, 34, 35, stand-in, 73 stub, 73

37, 41, 44–47 standard subdirectory, see also
aspirational goal, 11 de-facto, 18, 179 directory, 314, 315
classic, 308 standard output, 50 subroutine, 39
conference, 11 Startup class, 23, 27, 55, 63, subterfuge, 32
deterministic process, 125 76, 81, 83, 150, 173, 310, subtype, 227
pocket, 11 311, 313, 315 Subversion, 18, 178
practice, 182 constructor, 82 suffering, xxiv, 129
process, 177 Stash, 178 sum type, 160
science, 97 state sunk cost fallacy, see fallacy
traditional, 300, 308 application, 78, 121, 164 SuperFreakonomics (book),
SOLID principles, 300 change, 164, 165 132
sort order, 256 local, 165 supertype, 227
Danish, 256 consistent, 218 support agreement, 78
sorting algorithm, see illegal surgeon, 17
algorithm unrepresentable, 159 surgery, 17
source control system, see inspection, 230 survey
version control system invalid, 106–108, 159 geographical, 127, 128
spaghetti code, 261, 285, 319 mutation, 106 sustainability, 34–37, 40, 44,
special case, 212, 237 object, 106, 164 45, 47, 114
specialisation, 188 system, 249 versus speed, 67
Speculative Generality, 52 transformation, 88 SUT, see System Under Test
spelling error, 25 valid, 106, 144, 156 SUT Encapsulation Method,
split screen configuration, stateless class, 76, 80, 173 66
135 statement Swagger, 205
spoofing, 293, 294 formal, 41 Swiss Army knife, 158, 164
SQL, 78, 212, 238, 245, 299 statement completion, 281 switch keyword, 133
named parameter, 295 static code analysis, see code syntactic sugar, 143
script, 315 analysis system
SELECT, 123 static flow analysis, 144, 147 edge, 265, 266
SQL injection, 293, 295, 296, static keyword, 27, 67, 139, restore, 284
299 142, 147 running, 268
SQL Server, 78, 299 statistics, 178 System 1, 42, 43, 279
SSTable, 45 steering wheel, 41, 42 System 2, 42, 43
Stack Overflow, 14, 240, 251, stored procedure, 299 system tray, 277
280 Strangler, 210, 220 System Under Test, 66, 69,
stack trace, 309 class-level, 215–217 301, 316, 325
stakeholder method-level, 214 coupling to test, 107
Continuous Delivery, 276 strangler fig, 210, 211 description, 128, 304
disregard for engineering, STRIDE, 292–294, 300 sabotage, 233
31 string comparison, 255 state, 230
feedback from, 49, 85 stringly typed code, 58, 164, triangulation, 127
involvment, 308 242
meeting, 280 stroll, 239 T
prioritisation, 290 struct keyword, 207 tab, 315, 316
security, 292, 293 structural equality, 72 tagged union, see sum type
371
Index
take-off, 16 integration, 54, 208, test library, 58

tampering, 293–295 242–246, 297, 318 test method, 89, 119
task iteration, 96 add, 225
big, 276 manual, 82, 85 orchestration, 248, 249
complex, 16 non-deterministic, 246, test pilot, see pilot
getting started, 276 248, 250 test runner, 58
tautology parametrised, 89, 90, 107, Test Spy, see Test Double
assertion, 97, 233 300, 301 test suite, 53
TCP, 103 append test case, 119, build script, 55
TDD, see test-driven 225 execution time, 244
development compared to property, failing, 248
team 301 noise, 247, 248
change, 187 passing, 96, 97 safety net, 224
high-performing, 5 property-based, 53, 279, trust, 223, 248
low-performing, 5 301–305 Test Utility Method, 65,
team coupling, 308 refactoring, 301 324–326
regression, 241 test-driven development, 53,
team member
revisit, 116 63
new, 111, 150
slow, 243, 246, 249 acceptance, 61
TeamCity, 24
smoke, 82, 85 beginner, 119
technical debt, 7, 8, 178
state-based, 73 coaching, 191
technical expertise, 31
test case, 90, 91 enabling, 54
temperature, 326
append, 225 execution time, 244
terminal, 135
before and after, 245 mob programming, 316
terrain, 291
comprehensive, 304 one among alternative
test
exercise, 250 drivers, 76
acceptance, 61
good, 119 outside-in, 53, 61, 64, 68,
add to existing code base, redundant, 128 78
24 single, 88 poka-yoke, 159
as measurment, 127 test code, 224 scientific method, 97, 98
automated change, 234 security feature, 251
as driver, 53 coupled to production success story, 240
as guidance, 167 code, 228 teaching, 119
database, 83 duplication, 61 technology choice, 78
ease, 19 edit, 224, 225, 228, 232, triangulation, 127
favour, 82 234 Test-Driven Development By
system, 82 maintenance, 234, 325 Example (book), 116
boundary, 69, 76, 83 problem, 91 text file, 21
coverage, 118, 223, 250 refactoring, 227, 229, 231, textbook, 280
deterministic, 246, 250 232, 234, 326 The Leprechauns of
developer, 21 rotate, 56, 57 Software Engineering
example, 300 test data, 303 (book), 13
exploratory, 208, 241 Test Double, 73 The Pragmatic Programmer
failing, 63, 96, 241 Fake Object, 73, 84, 122 (book), 9, 279
high-level, 65 Test Spy, 229, 231 Theory attribute, see
in-process, 243 test framework, 90, 282 attribute
372
Index
thinking Data Transfer Object, 70 type system, 106, 157

deliberate, 42 input, 50 static, 28, 100
effortful, 43 Transformation Priority type-driven development, 53
thread, 57, 58, 76 Premise, 75, 89, 115, 119, TypeScript, xxiv
multi, 250 128 typist, 5, 281
race, 246, 247 tree, 151, 211, 314 typo, 187, 196, 281
single, 249 B, 45
thread safety, 80, 173 dead, 211 U
threat, 292, 299 fractal, 151, 152 ubiquitous language, 169
identification, 299 hollow, 211 unauthorised access, 293
mitigation, 295, 296, 298 host, 210 understanding, 235–238,
threat modelling, 292, 293, leaf node, 151 241, 251, 252
299, 300 LSM, 45 bug, 40
threshold, 130–133, 135, Trelford, Phil, 158 computer, 45, 176
306, 307 triangulation, 119, 123, 127 difficult, 35, 40, 46
aggressive, 154 geometry, 127 easier, 216
throughput, 326 troubleshooting, 235, 236, human, 45, 176
tick, 122 272 struggling, 182
time, 227, 264, 266, 273, 306 debugging, 256 undo, 18, 19, 186
management, 238 experience, 255 unintended consequence,
of day, 264 132
ordeal, 8
personal, 279 unit, 68, 69, 107
superpower, 256
wasting, 276, 279 Unit of Work, see design
support future, 272
time-boxing, 238, 276, 277, pattern
understanding, 268
280 unit test, see test
trunk, 151, 184
timeout, 248, 250 definition, 68
trust, 87, 224, 248
TODO comment, 67 universal conjecture, 98
try/catch, 92 urgency, 283
tool, 24, 72
TryParse method, 98, 99 Uri class, 58, 59, 66
analyser, 24–30, 53, 76,
88, 302 Twitter, 244, 277 URL, 66, 83, 205, 294, 296
warning, 28, 29 two-factor authentication, documented, 205
GUI, 82 see authentication opaque, 206
linter, 24 type template, 66
topology, 21 anonymous, 64 UrtCop, 26
Tornhill, Adam, 305 custom, 170 USB, 159
touch type, 280, 281 generic, 216 Usenet, 280
tradition, 9, 10, 268 polymorphic, 229 user, 4, 52, 296
traffic, 293, 298, 326 static, 164 regular, 293, 299
transaction, 87, 183, 246, wrapper, 302 user code, 151, 298
249, 250 type declaration, 28 user group, 31
roll back, 246 type hierarchy, 227 user interface
TransactionScope class, 250 type inference, xxvi before database, 6
transformation type information, 157 feature flag, 209
atomic, 88 static, 170 slice, 50
code, 88, 89, 91, 115, 119 type signature, 162 using directive, 21
373
Index
V vigilance, 320 wizard, 20, 54

vacation, 41, 187, 191, 205 vine, 210, 211 work
validation, 77, 92, 147, 154, violence, 210 design, 6, 278
261 virtual machine, 21 detective, 106
email address, 102 Visitor, see design pattern human, 13
input, 115 Visual Basic, 44 intellectual, 5, 42, 279
object-oriented, 144 property, 301 physical, 279
validation link, 102 Visual SourceSafe, 183 project, 4
validity, 99, 102, 106, 108, Visual Studio skilled, 8
147 add null check, 76, 81 uninterrupted, 276
value, 36, 37, 192 auto-generated code, unplanned, 192, 193
hard-coded, 303 21–23, 25, 72 work from home, 284
run-time, 273 build configuration, 25 work item, 275, 276, 283
Value Object, see design code metrics, 105, 133 work item management, 178
pattern, 72 developer, 30 workaround, 207, 255
value type, 207 generate constructor, 72 worker, 5
var keyword, xxvi, xxvii generate Equals and
Working Effectively with
variable, 75, 88, 89, 119 GetHashCode, 72
Legacy Code (book), 113,
count, 153, 154 Go To Definition, 315
223
global, xxv, 43 IntelliSense, 157
workshop, 292
local, 153 project, 30, 54, 318
worse is better, 36, 37
name, 196 solution, 30, 54, 245
wrapper, 207, 242, 269, 302,
VBScript, 44 test runner, 58
304
vendor, 78 void keyword, 162, 165
writer
version VT100, 135
single-thread, 249
language, 282 vulnerability, 293, 294, 296,
WYSIATI, see what you see
major, 219, 221 299
is all there is
new, 282
old, 282 W
X
platform, 282 wait time, 193
X out names, 162–164, 260
skip, 282 maximum, 194
walking, 239, 277, 279 x-ray, 307
version control data, 305
Walking Skeleton, 20, 54, 60 X.509 certificate, see
version control system, 18,
warnings as errors, 25, 26, certificate
19, 178, 305
centralised, 18, 19, 182 29–32 XML, 250, 326
CVS, 18 as driver, 53, 57 XP, 276, 284
distributed, 18, 186 cost, 67 xp_cmdshell, 299
secrets, 82 weak code ownership, see xUnit Test Patterns (book),
Subversion, 18 code ownership 224
tactical advantage, 186 web site, 51 xUnit.net, 55, 90, 245, 301
vertex, 314 Weinberg, Gerald M., 289
vicious circle, 193 what you see is all there is, Y
Vietnam, 6 43, 45, 152, 175 Yoder, Joseph, 153
view while keyword, 133
high-level, 151 Windows, 19, 22, 268, 277, Z
materialised, 299 315 zero, 102, 106, 107
374
Index
zero bugs, 240 zoom, 148, 149, 151, 152, example, 175, 265
zero tolerance, 25 154 navigation, 314, 317,
zone, see also flow, 42, 277, context, 314, 317, 324, 325
278 325 out, 265
375

Code That Fits in Your Head

Uploaded by

Copyright:

Available Formats

Code That Fits in Your Head

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Code That Fits in Your Head

Uploaded by

Copyright:

Available Formats

Praise for

Code That Fits in Your Head

Boston • Columbus • New York • San Francisco • Amsterdam • Cape Town

Series Editor Foreword xix

Chapter 1 Art or Science? 3

Chapter 3 Tackling Complexity 33

Chapter 4 Vertical Slice 49

4.3.4 DTO and Domain Model 70

Chapter 6 Triangulation 111

Chapter 7 Decomposition 129

Chapter 8 API Design 155

Chapter 9 Teamwork 177

9.2.1 Pair Programming 189

PART II Sustainability 201

Chapter 10 Augmenting Code 203

Chapter 11 Editing Unit Tests 223

Chapter 12 Troubleshooting 235

12.3 Bisection 250

Chapter 13 Separation of Concerns 257

Chapter 14 Rhythm 275

Chapter 15 The Usual Suspects 287

15.2.5 Information Disclosure 296

Chapter 16 Tour 309

Appendix A List of Practices 329

A.15 Parse, Don’t Validate 334

My grandson is learning to code.

Yes, you read that right. My 18-year-old grandson is learning to program

Anyway, my daughter asked me to spend an hour with my grandson teaching

to multiply numbers. Indeed, the PDP-8 didn’t even have a subtract

As I ﬁnished up the coding example, it occurred to me that I was scaring my

He and I disagree on all kinds of things. We disagree on static versus dynamic

He covers things such as Checklists, TDD, Command Query Separation, Git,

Moreover, there are gems scattered literally everywhere throughout these

Two years later, I was the proud author of a published book.

The core audience is ‘enterprise developers’—particularly back-end

You should be comfortable reading code in a compiled, object-oriented

A Note for S oftwa re Architects

You’ll ﬁnd no content about Architecture Tradeoff Analysis Method (ATAM),

The narrative is structured around a sample code base that implements a

Abou t th e Code Sty l e

I want the code to be comprehensible to as many readers as possible. Just as

For example, listing 2.1 includes this relative path: Restaurant/f729ed9/

Apart from family, my ﬁrst thanks go to my invaluable long-time friend

Mark Seemann is a bad economist who’s

This chapter presents some of them.

Try to understand what’s going on.

If you don’t understand why something doesn’t work1 , then make

12.1.1 S cien tific Method

Needless to say, this isn’t an effective way to deal with problems.

Your ﬁrst reaction to a problem should be to understand why it’s happening.

• Make a prediction. This is called a hypothesis.

1. Or, if you don’t understand why something does work.

The difference between this technique and ‘programming by coincidence’ is

• People develop complex Dependency Injection Containers [25] instead of

• People use advanced diff tools instead of merging more frequently.

To be fair, coming up with a simpler solution is hard. For example, it took me

You may be staring at a problem with no clue as to how to proceed. As the

2. Keep It Simple, Stupid.

The mere act of explaining a problem tends to produce new insight.

It doesn’t really have to be a rubber duck, but the technique is known as