Thanks to all who came along and contributed.

Summary

Attendees were asked, working in pairs, to recall development practices they had used on successful development projects, a successful project being one that:

• delivered the system it was intended to, a system than has been either
• sold to customers, or
• used voluntarily by at least 100 people
• took more than 3 months to develop
• had more than one person working on it

These practices were collated and ranked by frequency. The attendees then split into two groups and chose ten practices from those identified that they would want to use on a project. Each group prepared a poster describing their "methodology"

The Practices

Attendees identified these practices, and we counted how many projects they had seen each practice used on.

• work to an interface 28
• refactoring ++ 20 (can anyone remember what the "++" signified?)
• on-the-job training 20
• code reuse 19
• willing and available cutomer 18
• coding standards 16
• daily to-do list 14
• on-going code review 14
• time-out 14
• user acceptance testing 14
• just say thanks 13
• design up-front 13
• integrated version control 13
• honest estimating 13
• separate test team 13
• have a contingency plan 11
• change control 11
• frequent and regular builds 9
• reverse engineering 8
• formal code review 7
• automated regression tests 7
• manual regression tests 7
• manager takes the heat 7
• stand-up meetings 7
• architecture == business model 6
• non-integrated version control 5
• restricted email use 4
• personal development 4
• multiple estimation techniques 1

The "Methodologies"

4-4-2

• willing and available cutomer
• honest estimating
• just say thanks
• stand-up meetings
• design up-front
• code reuse
• work to an interface
• ongoing code review
• change control

(one missing)

Productivity ++

• willing and available cutomer
• honest estimating
• just say thanks
• personal development
• on-the-job training
• frequent and regular builds
• time-out
• version control
• manager takes the heat
• refactoring

Observations

• the two "methodolgies" were quite similar in some respects, but had differences in emphasis, one being concerned with specific tasks and techniques, and one with building a good working environment.
• The notion of using multiple estimation techniques when planning is a brilliant one. If the contributor who came up with that (from Philips?) could expand on how that works, I'd appreciate it. Thanks, Keith.
• your observation here...