Communications of the ACM

make sure there is always a pipeline of new ideas. You have to make sure the pipelines are set up so they never stall— keeping things innovative.

Kate’s takeaway: Think about the people around you. Do you have enough different opinions to keep your team out of a local optimum? How can you get more diversity?

What are the common mistakes that
even good engineering managers some-
times make?

Your prior assumptions are not higher order—that is, you are not making assumptions about your assumptions. I keep reading How to Win Friends … because I understand it is easy to fall back into my default behavior. That is the big thing: your work is never done. You never know and you aren’t perfect. There is always stuff to learn. You have to keep up with your trade and keep learning. You have to keep pushing yourself to get better.

Once you get stuck and stop pushing yourself, you are toast.

Kate’s takeaway: Think about some of the past lessons you’ve learned. What could you use a refresher on? What are some new things you want to learn?

What is your best interview question?

Given a generic type Cont r a = (a -> r) -> r, prove that this type forms a monad.

If you try to solve this question by brute force, you are going to fail. But if you look at it from the right level of abstraction, it is easy. So it forces you to problem solve your own problem-solving skills.

The particular formulation using
monads and type sounds really theo-
retical, but it is super practical. When
you are using JavaScript to write event
you make the system do things that no
one thought was possible?

Being agile is about communication. The process needs to change with the situation. You have to have a big picture of where you want to go, but any plan or process will shatter immediately when you hit your first bug or something happens out of your control.

In most projects there are two phases: an explorative phase and an execution phase. Your project should progress like a damped sine wave, where the amplitude gets smaller over time (see the accompanying figure). You have to figure out what to build, and figure out what question you are trying to answer. In the beginning you want to increase the vertical velocity to get uncertainty under control, and then you want horizontal velocity to increase when you get into execution.

With prescriptive processes, people are looking for a silver bullet to solve problems, but it doesn’t exist. It comes back again to “The Humble Programmer.” The world is super-confusing, and you have to embrace it and work with it.

Kate’s takeaway: You have to make your process work for you. Imagine your projects progressing on a damped sine wave—first focus on finding the right questions, and then the answers.

Who is the best manager you ever
worked for? What made this person
so great?

William Adams. He was my manager at Microsoft. He is an inspiration, and I am still trying to emulate him in my work.

There are several things I like about him. One is the importance he sees in diversity. For example, when dealing with feedback loops and prior assumptions, you need diversity to challenge your thinking. You have to actively put energy into creating a diverse environment so you are always challenging the status quo and maybe resetting your accumulated state. Don’t get stuck in a local optimum.

The other thing is that he always fo-
cused on people first. You want to create
the circumstances where everyone can
focus on their strengths. Always find the
best job for the person. Try to get a sense
of the progress and circumstances so
you can get ahead of what is next. For
example, if the project is winding down,
Being a great developer is difficult. It
requires constant learning and a pas-
sion for technology and science. The
same thing is true for great technical
leaders. There are a lot of smart les-
sons, but perhaps the most important
one is always to be pushing yourself,
and to be smart about your brainpower
and energy (working smart).

Related articles on queue.acm.org

Lean Software Development: Building and Shipping Two Versions

Kate Matsudaira http://queue.acm.org/detail.cfm?id=2841311 It Probably Works

Tyler McMullen http://queue.acm.org/detail.cfm?id=2855183 A Conversation with Erik Meijer and José Blakeley http://queue.acm.org/detail.cfm?id=1394137

References

1. Carnegie, D. How to Win Friends and Influence People. Simon and Schuster, 1936.

2. Colyer, A. The Morning Paper; https://blog.acolyer.org.

3. Dijkstra, E. W. The humble programmer. Commun. ACM 15, 10 (Oct. 1972), 859–866; http://dl.acm.org/ citation.cfm?id=361591.

4. Meijer, E., Kapoor, V. The responsive enterprise. acmqueue 12, 10 (2014); http://queue.acm.org/detail. cfm?id=2685692.

5. U. S. Marine Corps. Warfighting, 1997; http://www. marines.mil/Portals/59/Publications/MCDP%201%20 Warfighting.pdf.

6. Zuckerberg, M. Mark Zuckerberg’s letter to investors: ‘ The hacker way.’ Wired; https://www.wired. com/2012/02/zuck-letter/.

Erik Meijer has been working on “democratizing the cloud” for the past 15 years. He is known for his work on, among others, the Haskell, C#, Visual Basic, and Dart programming languages, as well as for his contributions to LINQ and the Rx (Reactive Extensions) framework.

Kate Matsudaira ( katemats.com) is the founder of her own company, Popforms. Previously she worked at Microsoft and Amazon as well as startups like Decide, Moz, and Delve Networks.