Video: Software, Faster - Dan North

April 13th, 2019

 

Video: Software, Faster - Dan North

Software, Faster - Dan North

https://www.youtube.com/watch?v=USc-yLHXNUg

 

 

 

 

patterns of effective delivery

 

 

Jeff Sutterland talks about hyper performing teams

from years to months

 

 

 

 

PATTERN:

repeatable strategy

depends on context

(vs best practice no context)

has

name

context (where it works)

warning (where it does not work)

 

 

best practice -> novice

 

 

design pattern book:

!= pattern

idioms to get by in a crappy language

 

 

EFFECTIVE:

goal - what are we optimising for

eg.

lead time

time to market

quality

learning

predictability

(agile = predictability)

what are we trading off?

 

 

DELIVERY:

"people don't want a quarter inch drill, they want a quarter inch hole"

 

 

 

 

===

 

 

 

 

Spike and Stabilize [development]

 

 

new code or stable code

no middleground code

not tested

not documented

not fresh: how did this work?

spike

experiment in code

metaphor: nail through wood asap

timeboxed

throw away

 

 

real options

options theory + real life

options

have value

expire

"never tie in early"

 

 

=> treat all code as a spike

// todo

review of shame

2nd time through

charactarisation test

'test driven testing'

 

 

only works if

-> you promise to stabilise

 

 

rules

second timeback

test

document

pay for cleanup

set timer

in 6 weeks

 

 

 

 

===

 

 

 

 

Hair Trigger [deployment/operations]

 

 

team deploys only code

can deploy quickly

can mess up quickly

 

 

safety catch

 

 

 

 

===

 

 

 

 

Ginger Cake

 

 

copying existing code

only code  you knowintimately

- structural

- behavioral

opposite of dry

 

 

 

 

DRY is a pattern, and has tradeoffs

dark side of DRY

coupling

 

 

 

 

===

 

 

 

 

Shallow Silos [team]

NOT "silo bad, cross functional good"

 

 

NOT rotation

shared standup

 

 

NOT everything or silos

 

 

 

 

===

 

 

 

 

Burn the Ships [learning]

 

 

I'm gonna build

this thing,

by this date,

using this technology

 

 

indirect learning

 

 

 

 

How lon

"On avarage between 5-15 years for full paradigm shift, because 3 things are needed

1. downward pressure (immovable deadline)

2. run out of options

3. information

So I make friends with companies and wait.

"

-- Eli Goldratt

 

 

 

 

takes a lot of discipline

"oh shiny, tutorials, look!"

 

 

no description for image available

 

 

no description for image available

 

 

leanpub.com/softwarefaster

 

 

 

 

dan@dannorth.com

 

 

no description for image available

 

 

 

 

 

 

Q&A

 

 

Eric Gamma ruined patterns?

 

 

Book: Fearless Change - Linda Rising

pattern language for organisational change

 

 

 

 

Why the name 'hairtrigger'?

sense of danger

gun, barely need to touch the trigger

 

 

 

 

How to sell the stabilise part?

traders

understand risk

 

 

 

 

Only seniors?

no, everyone