PI in the Sky – Paul Irvine

If not a dirty word, it certainly seems to have become a misused and abused word that represents many things to many people, and often has little – if anything – to do with the principles embodied in the Agile Manifesto. I’ve now had several CEO’s look to me for an ‘agile’ development team, and it is sometimes hard to really get to the bottom of what they as non-technologists, are really asking for.  Indeed, I’ve even had my own staff members coming from a legacy environment, looking to me for new and ‘agile’ ways of doing things.  I think they are all looking for something like ‘quick and dirty’ and not Agile at all.

I’m not sure that any CEO for any significant enterprise can really tolerate a full Agile approach.  Can you see the conversation with the board, and investors?  “Well, we can tell you exactly what’s coming this month, maybe what’s coming next month, but the big Summer release… not sure what will be in it yet. It depends what’s working in the iterations before it.  But whatever it is, we’ll be pretty confident we can sell it, and it will work!”

Yup, that’s going to go over real well, when budgeting for sales teams, marketing plans, trade shows, coordination with manufacturing, and the other mundane staffing and headcount things that all depend on what’s getting delivered on the roadmap. All those other segments of the business are planned over 12 to 18 months.  Revenue forecasts and expense forecasts are based on what that big “next Summer” release might actually contain.  How many developers on the team is determined by what projects are required to support those goals, and what those projects will deliver.  I believe in Excel terms, we’ve just hit a circular reference error.

I’m not suggesting that the other end of the spectrum – waterfall – is without issues either.  I’d represent it this way… Agile is akin to driving from Oregon to Florida. You know roughly which direction you are heading. South-East. You will see how far you get each day, and make judgments about which route to take at each major intersection based on current local info that you find, and your compass.  You only have  a vague notion of when you might get to Florida. The plus point is you’ve set out with minimal planning, and are making rapid progress from day one. The down side is software development teams are rarely working in isolation with the luxury of when they are supposed to turn up in Florida with the goods!

Waterfall is more like having planned the entire route before you set out, with turn by turn instructions printed out (remember those? it wasnt that long ago…). You have a reasonable certainty of when you expect to arrive. Indeed, everyone else is expecting you on that date, because that’s why you are heading there.  The objective is both date and destination. Your sister/brother/best friend is getting married perhaps!  The issue here is the inflexibility of the plan, and that it doesn’t take into account the local info along the way.  There is risk  that you might not make the date if the unexpected happens along the way. You may have to throw Aunt Maud out along the way, to gain a little speed or fuel efficiency!

I’m sure the happy trail is somewhere in-between. Most places I’ve been aren’t religious zealots about a methodology, thankfully. We’ve borrowed bits of XP, bits of Scrum, bits of Agile, and bits of Waterfall.   Balance in all things… what works for one team, may not work for another.  Somewhere between Waterfall and Agile we’ll find peace and harmony.  And software development will become a relaxing and fulfilling mental exercise.

Did you know that English is the only language where a double positive equals a negative?

Yeah, Right!