Monday, August 11, 2008
Incremental Death March
For those of you interested enough in Agile methodology and hold on to the hope that one day you will be able to implement this in your respective day job - I have news; "Culture eats methodolgy for breakfast". No, that's not my quote. I heard it from Ken Schwaber as I attended a scrum training seminar. And I'm not quite sure where he got it from. The prospect of having everybody participate in quality software development (Q/A, BA, Dev, Sales) sounds like a no-brainer. It's what eventually happens ANYWAY. Except instead of days, weeks, months or (dare I say it) even YEARS later - everyone who has a vestage interest in the outcome (or pigs/perhaps chickens) see each other everyday. You report your progress (or lack thereof to your respective co-workers) and "hopefully" everyone pitches in with their best effort to deliver on what the team decides is what is best for the product. Does it happen? Well sometimes, only under the best of circumstances - and with the RIGHT PEOPLE in place. Read Good To Great if you haven't already. Historically, management has gotten so used to dictating the what and when and the vast majority of developers have capitulated to such an extent that it is business-as-ususal. Cram in everything possible, to meet the DATE, letting quality slip as a direct consequence. The longer a "bug" lives (either unintentional or intentional - yes there are intentional bugs! These are bugs that are design-compromises made in the "heat" of a deliverable;"hopefully" with copiously scattered comments of TODO or FIX) With waterfall, I can give you software you don't want in 10 months, whereas with agile methodologies I can get you the same software in 1 month. At worst you lose 30 days as opposed to 9+ months. Is that good? Considering that "conservatively" 45% of program features are never utilized, it makes sense to take smaller bites of the most value than trying to swallow the elephant all-at-once. See Cargo Cult Methodology: How Agile Can Go Terribly, Terribly Wrong. Is software doomed? I think not. Takes a brass set to change and a lot of hard work. We're not there yet - but we're pushing the fly-wheel...
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment