"Count what is countable. Measure what is measureable. What is not measureable, make measureable." -- Galileo

Monday, September 24, 2007

Plone Metrics Goes Online

OK, so I couldn't talk management into Naples. That means I'll miss this year's World Plone Conference. And its an important one: Plone 3.0 is the current stable release and I need to spin up on all the new capabilities.

Best I can do for the Plone community is add my voice from afar. I've been corresponding intermittantly with Paul Everitt, this year's chair of the Plone Foundation, about metrics for Plone. One idea that's gotten a little traction is a method by Richard Bullocks from, of all things, the Air War College. He starts with a desired end-state and drills down to objectives, their values, and measurable attributes. See his dissertation for the full 188-page blow-by-blow description.

For Plone, I've come up with the following.

Plone End-State Objective Value Attribute
Stable yet evolving Software Releases Release Size
Release Frequency
Software Quality Bugs
Core mailing lists
Support mailing lists
Security vulnerabilities
Core Development Team Size of CDT
Involvement of CDT
Features New features
Widely adopted Acceptance Downloads
Installations
Defectors
Economic health of third-party companies
Technical reviews
Visibility High-profile installations
Users of Plone portals
Visitors to Plone portals


In a future posting, I'll add measures, sources of these metrics, and examine some of the ways we might obtain a valid and useful measure of Plone progress towards a successful end-state. Please comment freely on what you see as strengths and weaknesses of my methodology, not to mention on Plone itself.

No comments: