LEAN THINKING

  • Concepts
  • Patterns

TECHNICAL EXCELLENCE

LEAN THINKING

Concepts

  • % Utilization of Resources

  • Value-to-waste ratio

  • Cycle times

  • Value of information - also, create useful knowledge

  • Data Driven - experiment, collect data, adapt based on data

Patterns

  • Focus on early testing and feedback

  • Focus on large-scale test automation

  • Focus on continuous integration

Back to top

TECHNICAL EXCELLENCE

Architecture & Design

Concepts

  • Source code is the real design

  • Risk-driven development

  • Agility through low coupling

Patterns

  • Very early, develop a walking skeleton with tracer code

  • Incrementally build ‘vertical’ architectural slices of customer-centric features

  • Do customer-centric features with major architectural impact first

  • HTML-ize and hyperlink your entire source code, daily

  • Use lots of stubs, plus dependency injection

  • Use architectural and design patterns

  • Wrap calls to remote components with proxies or adapters

Issues

  • Intended architecture (speculated, wished for) versus actual architecture

  • Architecture is performed once near the start, often in documents

  • Architecture overlaps with requirements analysis

  • Start with indirect interaction between major components, then replace as needed

  • Message-oriented middleware (MOM) can be slow, may need to change to SOAP or Java RMI for remote clients.

Back to top

Test Driven Development (TDD)

Patterns

Short cycles of: 1. Write one test. 2. Implement just enough code to make it pass. 3. Refactor the code so it is clean.

Concepts

  • Short

  • Rhythmic

  • Incremental

  • Design-focused

  • Disciplined

  • Encourages lower coupling and simple, flexible configuration

Issues

When a developer creates a new component (such as a class) with TDD, or refactors a legacy component to be unit-testable, they must break the dependencies of that component so that it is testable in isolation

Back to top

Specification by Example or Acceptance test-driven development (A-TDD)

Patterns

  • Condense workflow in business rules

  • Expressing business rules in tables makes them more comprehensible

  • Use an A-TDD Tool

Back to top

Test Automation

Concepts

  • Scripted testing - automated

  • Exploratory testing - not automated

Patterns

  • Create maintainable tests
    • remove duplication in and between tests
    • delete tests when not adding value
    • do not test through the UI
    • run tests frequently

  • Treat non-functionals the same as functionals

  • Continuously run long-running tests

  • Use virtualization or containers

  • Avoid using commercial test tools

Back to top

Continuous Integration

Concepts

  • Shared Code

  • Working in small batches–iteratively.

Patterns

  • Integrating at least daily

  • supported by a CI system

  • with lots of automated tests

  • build and test need to be fully automated

  • Multi-stage CI:

    • a developer build
    • component or feature focus
    • automatic or manual promotion
    • event or time triggers
    • the number of stages

Issues

Many large products have legacy code without automated tests. Developers need to add automated tests–which is a lot of work.

Probability of breaking the build increases with more people checking in code. Address by:

  • build incrementally
  • deploy incrementally
  • manage dependencies
  • refactor tests

Increase in code size leads to a slower build and thus a slower CI feedback loop. Upgrading tools to the latest version or replacing a slow tool with a fast one speeds up the build a lot.

Back to top

Acceptance Testing

Concepts

  • UAT is a subset of acceptance tests

Back to top

Unit Testing

Concepts

  • Unit testing Facilitates changes

  • Unit testing Simplifies integration

  • Unit testing Provides documentation

  • Unit testing Serves as design tool

  • Total cost of ownership is low

  • Tests Internal Quality vs. External Quality

  • Feedback is precise and actionable

Patterns

  • unit test should be small and test only limited piece of code functionality

  • AAA: Arrange, Act and Assert

  • Behaviour Driven Development (BDD)

Back to top

Product Backlog

Concepts

  • Multiple teams building a single product work from a single Product Backlog.

  • Avoid unnecessary and costly complication by using the simplest tools possible for managing the Product Backlog.

Patterns

A good Product Backlog must:

  • have estimates for all items,
  • have finer grained items at the top and coarser grained items further down, and
  • be prioritized.

Back to top

Product Backlog Refinement

Patterns

  • split big items

  • do very lightweight item analysis for basic understanding

  • estimate items

  • identify strongly-related items that suggest shared work, common work, or coordination

Back to top

Coordination & Integration

Patterns

Communicate in Code - use continuous integration to keep track of what others are doing. When you synchronize, look at the changes others have made.