“i come to bury agile, not to praise it”

©Alistair Cockburn 2009

““I Come to Bury I Come to Bury Agile, Agile, Not to Praise It”Not to Praise It”

Effective Software Development in the 21st Century

Alistair Cockburnhttp://Alistair.Cockburn.us


2

I come to bury Agile, not to praise it;The evil methods do lives after them,

The good is oft interred with their bones, So let it be with Agile.

The noble Waterfall hath told you Agile was ambitious: If it were so, it was a grievous fault,

And grievously hath Agile answered it.

Here, under leave of Waterfall, Lean and the rest - For Lean is an honourable method, So are they all; all honourable methods,Come I to speak in Agile's funeral.

(based on Shakespeare’s Julius Caesar, Act 3. Scene 2)


3

Agile was my friend, faithful and just to me:But Waterfall says he was ambitious; And Waterfall is an honourable method.

Agile hath brought many projects home to Rome,Whose Payoff did the corporate coffers fill.Did this in Agile seem ambitious?

When the programmers have cried, Agile hath wept. Ambition should be made of sterner stuff.Yet Waterfall says he was ambitious, And Waterfall is an honorable method.

You all did see that in 2003 We twice presented him a kingly crown,Which he did twice refuse. Was this ambition? Yet Waterfall says he was ambitious, And sure he is an honorable method.



4

I speak not to disprove what Waterfall spoke,

But here I am to speak what I do know.

You all did love Agile once, not without cause. What cause withholds you then to mourn for it?



5

Bear with me.

My heart is in the coffin there with Agile,

And I must pause till it come back to me.



6

Developing software

consists of People

making Ideas concrete

in an Economic context


7

inventing, deciding, communicating, Solving a problem they don't yet understand

which keeps changing

Creating a solution they don't yet understand which keeps changing

Expressing ideas in languages they don’t understand which keep changing

To an interpreter unforgiving of error

Making decisions with limited resources and every choice has economic consequences

People


8

Software development is aCooperative Game


9

Infinite Organization SurvivalCareer Management

Competitive Cooperative

Open-ended Jazz music

Finitegoal-directed

King-of-the-hillwrestling

Poker

Tennis

Chess

Product Line Management

Rock-ClimbingTheater

JournalismExplorationSoftware

Development

Swamp game

IT Systems

Games have positions, moves, strategies


10

Two conflicting subgoals:

Deliver this softwareSet up for the next game

Three moves:InventDecideCommunicate

and the situations (almost) never repeat!


11

C6 C20 C40 C100

D6 D20 D40 D100

E6 E20 E40 E100

L6 L20 L40 L100

Project Classification Scale:

Number of people coordinated

1 - 6 - 20 - 40 - 100

Comfort

Essentialmoneys

Life

Discretionarymoneys

Criticality

X

X X

XX

Adapt to your situation


12

Richness of communication channel

Com

mun

icat

ion

Eff

ectiv

enes

s

2 people on phone

2 peopleon chat

(Questio

n-and-Answer)

2 people atwhiteboard

(Courtesy of Thoughtworks, inc.)

Face-to-face is the most effective - Try Video

Paper(No Question-Answer)

Videotape


13

Kim Pat12 people:= $100,000 / yr penalty

12 people = $300,000 / yr penaltyKim

Pat

Distance is expensive

Kim Pat Programming in pairs

“Managing the Flow of Technology” “Distance Matters”


14

People issues determine a project’s speed

Can they easily detect something needs attention?

(Good at Looking Around)

Will they care enough to do something about it?

(Pride-in-work; Amicability)

Can they effectively pass along the information?

(Proximity; face-to-face)


15

Software development is a Craft


16

Craft teaches us to pay attention to our skills and to the medium

Major crafts:

1 Deciding what to build

2 Managing (people and projects)

3 Modeling

4 Designing the external view

5 Large-scale design (architecting)

6 Fine-scale design (programming)

7 Validating the work


17

People learn skills in 3 stages

Shu: Learn a technique

Ha: Collect techniques

Ri: Invent / blend techniques


18

Use Lean Processes


19

I wish they’d decide what style they want!

Users &Sponsors

Testers

BusinessAnalysts

I wish they’d decide what

functions they want!

A decisionI wish they’d decide

on the UI designalready!

I wish they’d think a bit more carefully

before coding!

Programmers

Every line of

code is a decision

UI Designers

Software development looks like manufacturing if the unit of inventory is the

unvalidated decision!


20

Software development has correction loops

Detaileddecisions about

externalappearance

Decisions about function

and style

Detailed decisionsabout function

and data

Decisions about program

structure

Decisions about systemcorrectness

Users &Sponsors

UI Designers

ProgrammersTesters

BusinessAnalysts


21

From Lean Manufacturing: aim for continuous flow

time

Testers100 lines of codeshow up for testing

# of design decisions

Decisions being validated or broken

How long this line of code / screen / use case / decisionsits in the testing queue

Decisions arriving


22

From Lean Manufacturing: watch your queues

Enough Users & Sponsors

Many Business Analysts

Not enough Programmers Some

Testers

Users & Sponsors

absent

UI Designers

Business Analysts

some Programmers

Database designers

Tester

Enough Users & Sponsors

UI Designer

A few Business Analysts

Lots of Programmers

Single, solitary Databasedesigner


23

Design is Knowledge Acquisition


24

Delivers nearly no knowledge(or risk reduction)

Knowledge comes atthe “moment of truth”: final integration.

Waterfall is a late-learning strategy

time

cost

Growth of knowledge with big-bang integration


25

Development sequence indifferent (with respectto knowledge)

Delivers knowledge

(risk reduction)

We can pay to learn early in the project

time

cost

Growth of knowledge with early, continuous integration


26

Develop for business value once risks are down

time

Knowledge growing(risk reduction)

cost

Growth of business value

Business value growing


27

Trim to deliver on-time (or early)

Delay to get moreor better

Trim the Tail: Choose to deliver by value or date


28

In the 21st century, software engineering will use craft, cooperative game & lean principles

Craft developing skills in a mediumshu - ha - ri

Cooperative game of invention and communication teamwork, communication, strategies

Lean processes (“unvalidated decisions = inventory”) small queues, cross-trained people, varyied processes

Design as knowledge acquisition early integrationpay to learn trim the tail


29

http://Alistair.Cockburn.us

“i come to bury agile, not to praise it”

Documents