The Cathedral and the Bazaar

The connective tissue of the internet is built with free and open source software. Data centers run various distributions of Linux. You can browse the web with Mozilla Firefox. It’s highly likely your company stores data in one of many open source databases such as Postgres and MySQL. Maybe you have Lucene indexes served up in a distributed cluster by Solr. When you reflect on what we can download and use for free while also receiving continual updates from skilled engineers that contribute for a hobby, it’s quite staggering.

As of 2018, the hosting site for software projects using the Git version-control system, GitHub, had received over 1.1 billion code contributions from 21 million developers across 96 million projects. Popular open source projects such as TensorFlow, a machine learning framework, have accepted contributions from over 9,000 developers. The React framework for JavaScript has over 10,000 developer contributions. If you were beginning to get into software development today, you may think this is the way open source software has always been done: out in the open with hundreds, if not thousands, of contributors.

However, the watershed moment for this model was the development of the Linux operating system. The Cathedral and the Bazaar [Ray01] is an expansion of an essay of the same name that drew attention to how radical it was at the time. Before Linux, many open source software projects were “carefully crafted by individual wizards or small bands of mages working in splendid isolation.”

The title of the book comes from the two different software-development models that it covers:

• The cathedral model, in which software is free and open source but is developed by the aforementioned hermit wizards. In the book, the GCC compiler is given as an example. These projects have long release intervals and are controlled tightly by the development team.

• The bazaar model, in which the entire world contributes and opens their development to the public, despite the chaos. They operate under the tenet that “given enough eyeballs, all bugs are shallow.” There’s less to lose if occasional issues slip out. Linux is described as the poster child for this movement.

Now, step back from this a second. What has this got to do with management? Well, let’s find out. Let’s consider the potential motivations of engineers to want to work under both models.

• In the cathedral model, we could reason that the maintainers of the project are most comfortable when they have control: that is, they work in a small cabal of level-20 wizards that know each other well, have a final say in their own destiny, and are subject to less distraction and noise from the outside community while they are doing so.

• In the bazaar model, the project maintainers are motivated by change: that is, they prefer the fast pace of development by hundreds, if not thousands, of contributors and the fact that all bugs are easy to solve if a large number of people are working on fixing them.

Now, we’re not here to debate the relative pros and cons for each model. However, I think that the models highlight relevant individual personality traits and motivations that are well worth thinking about as a manager—that is, that some individuals are highly motivated by control and stability and others are motivated by chaos and change.

Think of a real cathedral. It is beautifully architected using the finest material and craftsmanship. It is meant to stand the test of time, both in terms of use and the number of years that the building is expected to be on this planet. It evokes quiet contemplation and peaceful reflection. Now think of a bazaar. It is heady with aroma and the voices of sellers. It’s bustling and ever-changing. Within the chaos are incredible new things to be found: exotic food and beautiful garments. Those that are driven by new experiences are drawn to the bazaar to be swept away by the experience.

Think of your team for a moment. Given that you’ve worked with them for a short while, have you begun to see which of them are cathedral constructors and which are bazaar browsers? How would you categorize yourself? The important consideration is that regardless of whether your staff thrive within chaos or stability, neither is right or wrong. They are just personal preferences and motivations; however, they may change over time depending on a person’s life, circumstances, and drive.

The challenge for you as a manager is to have the conversation: identify which of your staff are in either camp, and then build that into how you delegate to them, the opportunities you give them, and the environment within which they work. Let’s have a look at both types of people in turn and give some example situations in which they’ll thrive. Bear in mind that those who index strongly to either end of the scale may be harder to integrate into teams and may require careful management. That may include moving them to a different team.

Cathedral Constructors

Your cathedral constructors yearn for focus, stability, purism, and craftsmanship. As a manager, you need to be able to offer them the following opportunities:

• To become subject-matter experts. Is there the opportunity for them to build and contribute to a core area of the technology stack? Can they be given ownership over it in such a way that lets them become the most knowledgeable in the company about that particular part?

• To be a cornerstone of the team. Is there an opportunity for them to become the individual that their teammates seek advice from when approaching something new? Are they able to provide the scaffolding that new functionality is built on top of?

• To go deep, not wide. Your cathedral constructors relish in thorough knowledge. Is there an opportunity for them to contribute back to open source projects that they are an expert in? Are they able to purposefully stay away from newness in order to practice deep mastery?

• To show others their own ways. Those that become experts also benefit (and often relish) in teaching others what they know. Ensure that your cathedral constructors have the opportunities for mentoring.

• To revel in the detail. Try to keep incomplete knowledge or politics away from your cathedral constructors so they can focus on the small details. Let them be in their happy place.

Bazaar Browsers

The bazaar browser wishes to sample the sights and sounds of as much as possible. They want excitement, variety, chaos, and change. This is where they thrive and grow. You need to be able to offer them the following opportunities:

• To get as much newness as possible. Is there a new project coming up? Your bazaar browsers want to sample new challenges, domains, products, and technologies as much as possible. Put whatever new thing you have in front of them first.

• To build and throw away. Are you about to go through a phase of quick prototyping and iteration? Are you going to be experimenting with some new technology? Your bazaar browsers love this. Send it their way.

• To avoid stagnation. Unlike the cathedral constructors, the bazaar browsers find working on the same area for a long time a sign of stagnation rather than mastery. Ensure that you’re talking with them about how they feel, and give them opportunities to switch up what they’re doing as time goes on.

• To even move around teams. If your team is unable to keep a bazaar browser motivated, then why not consider loaning them out to another team for a short while? They’ll get to do something new and fun and bring back knowledge. You could even take on a bazaar browser in return to inject some new energy into your team.

Having the Conversation

While many may have had a discussion with their manager about whether they want to become managers or not in the future, one could speculate that far fewer have discussed whether they are cathedral builders or bazaar browsers. This is great material for your one-to-ones.

You may also find that your preconceptions about some of your staff could be completely wrong: maybe Bob was underperforming recently because he needs more variety, but you thought it was best for him to stick with what he knows. That was wrong! Maybe Alice is happiest when she can stay inside her comfort zone, and the new challenge you gave her is uncomfortable rather than exciting.

By knowing what drives people, you can better place them in upcoming projects and teams; both for their happiness and for the best performance for your business.