Software design is an exercise in human relationships. Yes, it has its technical constraints. Mastering those techniques to resolve those constraints will improve the resulting designs. But the design only learns at the pace of the people living within that design.

In this workshop we will examine software design through various lenses:

• Economic. Software design increases the option value of software.
• Social. Software design as community learning & change.
• Statistical. The curious prevalence & enormous implications of Pareto distributions in software.
• Practical. Making large changes through small, safe steps.
• Managerial. How to manage the change of a software design.

Exercises include:

• Tidying. The smallest changes of design.
• Refactoring. Changing design smoothly while maintaining a flow of new features.
• Code review & design. Separating commits into pure behavior change & pure structure change.
• Analysis. Finding & understanding the implications of distributions in software.

Come prepared to program.

Genies are great at cranking out features. Unfortunately they don’t clean the kitchen as they cook. Progress slows then stops as yesterday’s expedient decision becomes today’s roadblock.

What would it take to get the trajectory in red instead of the one in blue?

• Development in small steps validated by automated tests
• Periodic pauses in feature development to improve structure and thus increase optionality
• Potentially a new approach to augmented coding manipulating the tree structure of the code, not just the tokens

Sound familiar?

Genies are great at cranking out features. Unfortunately they don’t clean the kitchen as they cook. Progress slows then stops as yesterday’s expedient decision becomes today’s roadblock.

What would it take to get the trajectory in red instead of the one in blue?

• Development in small steps validated by automated tests
• Periodic pauses in feature development to improve structure and thus increase optionality
• Potentially a new approach to augmented coding manipulating the tree structure of the code, not just the tokens

Sound familiar?

Genies are great at cranking out features. Unfortunately they don’t clean the kitchen as they cook. Progress slows then stops as yesterday’s expedient decision becomes today’s roadblock.

What would it take to get the trajectory in red instead of the one in blue?

• Development in small steps validated by automated tests
• Periodic pauses in feature development to improve structure and thus increase optionality
• Potentially a new approach to augmented coding manipulating the tree structure of the code, not just the tokens

Sound familiar?