Beyond Agile Software
1 like135 views
The document discusses advanced methodologies in agile software development, emphasizing the need for individual ownership and creativity within teams to drive innovation. It critiques conventional agile processes for causing stagnation and advocates for a flexible approach that adapts to project complexity, using concepts like hyper-agility and continuous improvement. Ultimately, it calls for a balance between established processes and individual initiative to maximize value delivery.
Beyond Agile Software
- 1. Beyond Agile Taking an Agile Software Development process to the next level
- 2. The Death of Agile More memorable title
- 3. Satan loves Agile Now you’ll remember this talk
- 4. Advanced Software Development Processes More accurate title
- 5. Gabe Hamilton What who am I? Jabberwocky superfly! 1 I’ve worked on lots of kinds of projects. Short, Multi-year, Green-field, Maintenance mode Government, Fortune 500, Startups Agile -> 2005
- 6. Agile Software Development Methodology Why?
- 8. But not this order After 40 years of being wrong we decided to try something different. https://www.slideshare.net/gabehamilton/soft ware-engineering-for-ceos
- 9. We did it! We have a good agile process! 1 Regular Retrospectives Regular planning meetings 1 week, 2 week, 4 week, whatever 2 Estimate Stories Prioritize !!! 3 Some pairing Or at least code review 4 Tests that we like Some tests we don’t like too but that’s life
- 10. Feature slip We deliver releases. Rather than write godawful code, we let features slip to the next iteration. Mostly.
- 11. It doesn’t matter what flavor 1 Scrum Let’s under/over commit each week 2 Kanban One thing at a time please. 3 Lean Let’s pretend to build the software 4 Extreme Programming Why are we yelling!
- 12. We are in lock step, delivering value Iteration 1 User Interface Iteration 2 Initial API 3 Dashboard 4 Single sign on 5 Message Bus
- 13. So what’s the problem? Assembly line Lack of individual ownership & creativity Rule of committee
- 14. We all agree to limit ourselves. And some of that is good. We do need common vision and alignment, and to all pull together in the same direction
- 15. Who here is a repetitive machine?
- 16. Our future robot overlords aren’t In Machine Learning we can use prediction error to train an algorithm. For example, training on video to predict traffic by time of day. Areas of maximum error are the ones to focus on (Curiosity). Except focusing on areas of error leads to things like trying to predict the color of the next car. “Boy I’m bad at this, better keep trying” So we program in, “Don’t keep trying if you aren’t getting better” (Boredom).
- 17. What are the negative effects? 1 Everyone quits 2 Focus on incremental value 3 Lack of innovation 4 Stuck in local maximum
- 18. So what do we do? No Product Managers? https://vimeo.com/74437772 Watch Tim’s excellent talk But wait, I like some Product Management...
- 19. So what do we do? Individuals and interactions over processes and tools Working software over comprehensive documentation Customer collaboration over contract negotiation Responding to change over following a plan Return to First Principles
- 20. So what do we do? Kaikaku Radical change: rearranging the assembly line over lunch. Key part of continuous improvement. We can radically change our agile processes
- 21. We can have both Agile Process is a tool We have many tools We can use different ones at different times
- 22. Hyper-Agile
- 23. Hyper-Agile Good before and after releases. Every day: What is the current blocker? What most needs to be fixed? Standup is the planning meeting.
- 24. Long form Good for building value 3 month projects Big, vague problem definition
- 25. Long form - full autonomy “Work on what you want... and demo it” People will solve what is bugging them. Often the most boring part of the project. And will create outsized value.
- 26. Adapt approach to current complexity https://www.slideshare.net/gabehamilton/adaptive-software-engineering For a large new project To get a team up and running and to bring on new members use a normal agile process for up to 3 months. To create maximum value we should spend the most time in long form. Once the major pieces are in place return to agile for a few weeks, then switch to hyper-agile for up to two weeks before and two weeks after full go live. Start Full go live
- 27. Half measures Hackathons Design sprints Useful techniques. Use these as phase transitions into long form projects.
- 28. True Agility = Spare Capacity Theory of Constraints: Protective capacity “Any part of a system that needs protection from uncertainty, variation, or disturbances in the environment, while still interacting with that environment, requires some sort of buffer.” -https://praxis.fortelabs.co/theory-of-constraints-104-balance-flow-not-capacity-60baa74ce9f4/
- 29. Questions?
- 30. Acknowledgements Smartphone road Photo by Matheus Bertelli from Pexels Robot Photo by Somchai Kongkamsri from Pexels 1. Lyric from Save Yourself byAesop Rock
- 31. “Squishing people is boring, it’s so predictable”
- 32. Thank you.