Came across Uncle Bob Martin’s Clean Agile talk1, putting words to paper on it… DRAFT so far
It seems a lot of the original folks from the original Agile Manifesto are speaking up about how “the industry " got it all wrong. Dave Thomas emphasizes, it’s a adjective not a noun. Martin Fowler, emphasizes that SAFE is Shitty Agile for Enterprises. I came across a presentation from Robert Martin introing his book Clean Agile recently too.
He says that to them I think, the ideas in the Agile Manifesto needed to be expressed because the dot com boom was a kind of inflection point where a whole lot of people got into the software industry , but they agreed that many people were stuck trying to waterfall their way through projects, but it wasn’t really working.
I have personally not yet been on a team that has figured it out yet . I gather from another thinker in the space, Allen Holub, that many teams assume there is a cookie cutter approach that will just work. But more or less he says you cant buy agile, you have to adopt the mindset. And mindset shifts are very difficult.
I may have read this or Im making it up, but I feel like most of the teams I have been on that dont “try” to do anything , were my favorite experiences. People just talk , they talk about the work, stakeholders ask uncomfortable questions, individual contributors like myself get anxious 😅, try to bury the lead, and figure out how to lay out the sh*t sandwich of the day. Eventually you get to talking about facts and through conversation, you just figure out how to steer the work. And thats it. Its uncomfortable but it works.
On the other hand, teams I been on that try to do agile, do the song and dance but it never really worked. It turns out that if you stand up and flap your wings like a bird, you will not fly!
In Uncle Bob’s talk, he conveys the idea that the purpose of agile and scrum and all of this, is to produce data. It’s the only purpose, to convey it more directly. He puts it even better, that Agile makes you lose all hope, but you trade it in for data!
A typical question on the team Im part of , lately is, should you split up a card if youre crossing over to another sprint. Before I wasnt sure, but given the data angle, I think splitting is fine, but as long as it is a deployable end to end unit.
So if you can cut scope out of a card, but still keep a deployment in there, that is a unit.
Perhaps the way you cut up cards may in the long run not even matter, becauss there is a kind of conservation of energy going on. Cutting a card arbitrarily doesnt hide the work remaining, unless you literally cut the scope.
And in aggregate , you should be able to see a pattern. But thats the other thing I learned what we were doing wrong on our team. We’re not burning projects down, were just burning a bonfire that keeps burning forever, as a platform, with no beginning or end. Its middle all the way through 😆.
Building the wrong thing
Uncle Bob mentions that if the team is building the wrong thing , thats got nothing to do with Agile. Thats management not doing their product research.
This makes sense to me. All of these Agile/Scrum/Extreme ideas, they are all just tools. They help you with , this arrow below.
A ---> B
But they dont define your “B”. …