I’ve been re-reading the classics of computer science (thanks MakerBot HR! book-fund) and The Mythcial Man Month was on the list. This post and some posts that follow will revisits the book, and add some of my own theories to it.
The title comes from the once (and again) myth that there is such thing as an interchangeable man-month (hacker month, HM for this post). Much of the books is predicated on a simplification that many businesses make (or need to make) to even pretend to make sense of development . Imagine you can take a person and abstract their skills into a ‘hacker-month’ of code. Lets call that 1HM as the base unit. 1HM = 1 month of l33t development.
So, before we get even into the depths of the code on this, lets start with what a horrible abstraction this is. I know a lot of developers, and I can tell you right now, talking about code productivity as a interchangeable format like that is just insane. Making code isn’t making widgets, they are not interchangeable. Like writing essays, style changes. You can have 100 authors write the same article on the Wall Street Mafia, and still have only one be the right article.
So, to even start talking about this book, one needs to start with an already dubious axiom. That you can make something approaching comparisons in development skills. But for the following posts on Mythical Man Month, lets buy those Axioms, and move forward knowing our theoretic cornerstone is shaky.
Dubious Axiom 1:
One can quantify coder ability on a liner scale, which we will call the HM (Hacker Month).
Whether you call this in lines of code per month, or functions per month, or whatever ‘per month’ you want to measure, imagine there is some kind of way to track productivity that translates across all kind of developers. Suspend your disbelief, at least for a bit.