Moving to flow-based softeate development An experience report by Jørn Ola Birkeland , Bekk Consultng. Report: LIU TENG YUAN. Background. The iteration S hort iteration (two weeks) generalist developers v.s domain experts. Fig 1: Day one of the iteration. Looking good .
Report: LIU TENG YUAN
Short iteration (two weeks)
generalist developers v.s domain experts
Fig 1: Day one of the iteration. Looking good
Fig 2: Last day of the iteration. Not exactly what we’d planned
Burn-down charts by ScrumMaster
Retrospective (valuable , but suggestions ?)
Fig 3. The burn-down at day one. Expected worst case best case and average case added. In this case the team had planned a little more than could be expected in the average case.
Fig 4: The actual burn-down after the iteration. The team did pretty well, but was far from delivering all that had been planned. Scope change is very visible.
The release cycle
three 2-week iterations( test week , stage week ,development)
test week: regression testing , user stories
stage week: more regression testing , no errors , final verification
1.test week was very stressful
2. 3rd deployment had to be carried out
idea: next release cycle started in the stage week of the previous one.
The revolution Time box/iteration was eliminated Stand-up meeting changed from team member to work item focus Support process and bug handling integrated in regular process Work item estimation skipped Burn-down charts dropped A defined workflow was introduced , with clear goals/deliverables in each step
The whiteboard was changed
WIP limits were introduced
Fig:6 Bugs reported in various environments for each release. FSD introduced after 20090508
“Moving to flow-based software development” by Jorn Ola Birland