Drupal creator, Acquia founder and open source advocate, Dries Buytaert shares the lessons he's learned over the past two decades as a coder, CXO and IT leader.
Many in the open source community shun all things commercial, but they shouldn't, says Dr. Dries Buytaert, the creator of Drupal and the co-founder and CTO of Acquia. In this episode of , I talk with Dries about this and other lessons he's learned during his 20 years as an open source leader and advocate. The following is a transcript of the interview, edited for readability. You can listen to the podcast player embedded in this article, watch a video above, or read a transcript of the interview below, edited for readability.
Drupal to Acquia: From personal project to open source business
Bill Detwiler: And it's exciting, not just for me to have you on the show, but also because it's the 20th anniversary, just a few days ago actually, in January, Drupal turned 20. That's two decades. Before we talk about some of the lessons you've learned in those last two decades, give me a little bit of a maybe timeline of Drupal's evolution over the 20 years, and maybe even just a little bit about your personal journey too. What's changed a lot?
Dries Buytaert: Yeah. Sure. Yeah. Well, I started Drupal when I was actually a student. I was studying computer science in Belgium, where I was living at the time. And as a student, I was living in a dorm. And for the purposes of my dorm, I wanted to build a website, an intranet really, so I could exchange messages with some of the other people living in the dorm. And at the time, I wanted to build this application myself as a learning experience for how to build websites, and PHP and MySQL at the time were sort of the new kids on the block, were like new inventions. And I'm like, "Let's use PHP and MySQL to build a intranet application/message board," and that's really how Drupal got started. It wasn't called Drupal at the time, but when I finished college and I moved out of the dorm, this little website was actually a lot of fun. It's how we'd stay in touch and share things, interesting links that we find on the internet, and I felt it was like a shame to shut it down as I moved out.
So I figured I would move my intranet application to the public web, and so obviously had to register a domain name for it. I moved it to drop.org, that was the name of the website at the time, and the internal message board actually started to evolve into more of an experimental platform, so I started adding new capabilities like blogging. And this was before blogging was called blogging. It was actually called public diaries, and added things like RSS at a time when RSS was kind of being invented. I would lurk and see what people were doing and I would add all of these capabilities to my little site. So slowly, but certainly, my website, drop.org, which was now on the public web, started to attract people interested in the future of the web because I was experimenting with all of these things.
And so eventually I decided to make the software behind my site available as open source, and I spent maybe 30 seconds thinking about a name, came up with Drupal, literally copied the GPL license file from my Linux kernel into my website, and created a tarball, or a zip file, if you will, and uploaded it to my site expecting maybe 10 people to use my software. And that's the day that Drupal was born, and that was on January 15th, 20 years ago. And ever since, actually, I've been working on Drupal. So I've been working on Drupal for 20 years.
SEE: Best open source innovations of the decade (free TechRepublic PDF)
For the first seven years or so, it was a hobby project. It's what I would do at night, on the weekends, basically every moment that I had. Every second of free time I was obsessively working on Drupal. And then after seven years, I decided to turn my passion, Drupal, into my full-time job and started a company called Acquia, which was born as a company around Drupal, an open source business, if you will, to provide products and services around Drupal.
But those first seven years, I was actually working as a software engineer on embedded systems. I also did a PhD in computer science, which was all about runtime compilers and garbage collectors, so nothing to do with Drupal, or nothing to do with the web, but that was a lot of fun too. The last 14 or so years, I've been full-time at Acquia, helping to grow Acquia, while continuing to work as a project lead for Drupal too.
Bill Detwiler: So what would you say is maybe the biggest transformation that you had to make going from someone who's developing this project, sort of personally and for your own... Sort of what you wanted it to do, your own needs, versus now running a multimillion dollar company that has employees, that now has a "product," that has to meet other people's needs, and you taking on a leadership role as the head of the company. How did you transition?
Dries Buytaert: Yeah. I had to learn this for both Drupal and Acquia. So just to put it in context, Drupal is one of the largest open source projects in the world. We power one out of 30 websites. We have over 8,000 individual contributors a year. We have over 1,200 organizational contributors a year, so I think that makes... We're one of the largest open source projects in terms of participation. Acquia itself has also grown to be over 1,000 employees, so we're no longer a tiny company. But anyway, to go back in time, as I mentioned, I released Drupal as open source expecting maybe 10 people to use it, but I really vividly remember, for the first, I don't even know, seven years or so of Drupal, I really felt like I was an accidental leader. I didn't really sign up to be a project lead.
But as Drupal started to grow, I really had to pivot what I was doing, because I started out as the only developer for Drupal, so I was writing codes full-time, to all of a sudden have to start caring about other contributors, people that wanted to help make Drupal better. So I kind of find myself, all of a sudden, in a position where I had to provide leadership. And at least at the time, where I went to school in Belgium, there was no leadership classes as part of the computer science curriculum. Maybe that has changed, but 20 years ago, that wasn't something that you learned. And that was a weird transformation, and a slow transformation.
Also, to learn how to deal with a spotlight, for lack of a better term. I'm sort of an introvert by nature, but all of a sudden I would be doing presentations in front of 3,000, 4,000 plus people, and that didn't come natural for me, to be on stage. I'd kind of talk about the future of the web, or the future of Drupal, like all of these things I had to learn, and I felt kind of forced to learn them because I wanted to do a good job at growing Drupal.
Open source shouldn't mean anti-commercial
Bill Detwiler: Yeah. I think the last time you and I spoke, we talked about that. We talked about that transformation from being an individual contributor or writing code every day to becoming a project lead, becoming a head of a company. I remember you talking about you sort of absorbing as much information as you could, surrounding yourself with people who were good leaders, learning from people that you saw did a good job with leaders, and it's a shame that if it's not being... I haven't looked at the computer science, engineering curriculums at a lot of colleges around the world lately, but it is a shame that it isn't taught. If it isn't being taught, it should be, because you think in the MBA programs, you think in some of those management programs, that there might be those leadership courses, or at least they might be components of other courses.
But myself, I remember starting out as an engineering major in college. It was either mathematics, or it was a little bit of code, and that was kind of it. And then you got more code, more... But there was nothing in the terms of... Maybe they would throw you on a team and see how you did, but there were no lessons to help you either be a team contributor, or even lead. So that's definitely something I hope people take away from your story.
SEE: Open source vs. proprietary software: Pros and cons (TechRepublic Premium)
Let's shift gears a little bit. Let's talk about how open source itself has changed in the last two decades, because I think that really helps influence those kind of lessons that you've learned in the last two decades, that things that you didn't know back then. And I think before the call, one of the first things... One of the things you and I were talking about that we wanted to discuss during the interview was something about how you... And you realized this early, which is you needed to find a way to monetize Drupal to incentivize cooperation. Talk about that as a lesson.
Dries Buytaert: Yeah. Sure. Yeah. I think in the early days of open source, certainly when I started 20 years ago, it was a little bit of a renegade movement. It was sort of commercial influence or participation was kind of like... Wasn't part of the plan. In fact, open source was very much going to be antiestablishment in a way, right? And also, open source wasn't trusted by enterprises or organizations either. They wouldn't believe that a group of volunteers could build software that was secure, stable, scalable, performance. So there was this kind of weird reaction in both directions, I would say. But over the last 20 years, obviously, a lot of that changed, where open source is accepted. It's probably hard to find a single company that doesn't use any open source in their stack. And also, when I started, there was maybe a handful, or a dozen or so, open source companies. Like open source as a business model wasn't proven either. I remember the largest companies were pretty small at the time.
And I think over the last 20 years, obviously, open source is a lot more accepted. Open source business models have matured. It's still rare, an open source business, compared to the hundreds of thousands of proprietary technology businesses, so I'm pleased that things changed quite a bit. Now on the commercial interest, I think what we've learned is that maybe in the early days that was a mistake, to shun away from commercial involvement. What we've learned through Drupal, for example, is that today, actually two-thirds of all of the contribution through Drupal is actually sponsored by an organization, which is a lot, right? So that suggests that commercial involvement, or that sponsorship of individual contributors, is pretty important for the scalability, the sustainability of an open source project. And now, actually believe we need to embrace it, promote it, and really encourage organizations to contribute back to open source.
And another thing that has changed is, I think, people are starting to realize that volunteer contribution is kind of a privilege, too. Right? Not everybody has the time to contribute, as an example, so that's something that I would say we didn't really think about 20 years ago when we talked about open source. It's one of the reasons why, for example, we see less contribution from women, because very often, still today, women have to juggle a full-time job, have to still do more work in the households, as an example, so that leaves them less time to contribute on a volunteer basis. So I think there's a lot of reasons. Scalability, diversity, all of these are good reasons, I think, to embrace commercial involvement with open source. And as open source grows up, we're starting to realize that commercial involvement isn't bad after all.
Bill Detwiler: How do you balance that? Because there were those purists, especially back then, who wanted to keep... Like you said, that wanted to keep commercial involvement out of open source, but then it does run up into the roadblock of I've got to eat, I've got to pay my rent, I've got to pay my mortgage, I've got to survive. So balancing the two, and you still need income for all those other things, until we get some universal basic income, or everyone just doesn't charge for anything and everything is free, and the incentives, those monetary incentives, kind of go away, if that ever happens. But was there a tipping point, or how do you even now talk to people about, who are purists, on either side of that equation?
SEE: How to build a successful developer career (free TechRepublic PDF)
Dries Buytaert: Yeah. I'm not sure there's been a tipping point for us. It's been a very gradual increase of commercial involvement every year. We have more commercial contributors, if you will, and less volunteer contributors. But just to be very clear, volunteer contributors play a very big role in Drupal's development, and in opensource development in general, so that's very valuable, but I'm not sure we have to balance it. What I think we have to do is have good governance, meaning have decision-making algorithms for how we decide on features and functionality, what goes into the next version, how do we make certain architectural decisions, how do we determine membership of our community. All of these, we need good governance around.
And then whether the participants are volunteer participants or sponsored participants, I don't think it matters that much, as long as you can have a system of a governance model that takes input from all the stakeholders and balances the different interests in the project. You know what I mean? It's less about who contributes and how do they get paid, or their contribution time. It's more about how do you manage all of the different stakeholders, and how do you have a good governance mechanism that takes into account all of the needs of the project, those of volunteers, as well as those of commercial interests.
Diversity and inclusion are critical for open source to succeed
Bill Detwiler: Right. Now you mentioned something, too, that I think I'd love to hear more about, which was trying to include more... And kind of flows from that, which is trying to include more people in open source development in general, whether it's Drupal or another project, and that's something that's changed dramatically in the last two decades. Talk a little bit about how diversity, how increasing inclusion, has changed.
Dries Buytaert: Yeah. Well, first of all, open source projects have done really poorly, I would say, in terms of diversity and inclusion compared to maybe regular technology companies, who also do poorly, by the way, but maybe open source has done worse historically. I think it goes back to some of the things that I mentioned, where free time is a privilege, and therefore not everybody can contribute. But anyway, I think it's really important, diversity and inclusion, because that's how you build the best software, when you get a diversity of opinions and inputs and ideas. It's how you build the best software. And for Drupal in particular, it's really important, because as a content management system that powers one out of 30 websites in the world, what that really means is that if anyone visits more than 30 websites, let's say, statistically, they're going to hit a Drupal site. Right?
So we're really building Drupal for everyone, because I would say most people visit more than 30 websites. So everybody at some point is going to hit a Drupal site. And when you're building software that is used between codes by everyone, you need to make sure it's built by everyone, that you get that representation, meaning people in North America, but also people in countries or regions where maybe internet is slow, where they speak all sorts of different languages than English. How do we make sure that we build software that everybody can use, regardless of language, location, internet speed, all of these things. And the best way to achieve that is not for us in North America to think about how we can do that, but to actually involve and enroll those people into the development of our software, if they're able to and willing to, obviously.
So the diversity just helps us build better software, software that can be used by more people, and we have tremendous success stories. For example, we have been focused on accessibility for a long time, but when we saw that... At different points in Drupal's history we've had visually impaired people, blind people, contribute, become active participants in the development of the software. And it was amazing to see how they were able to impact and influence our accessibility features, because they are themselves blind, they could really help us understand how to make Drupal better, and we listened to those people and incorporated a lot of their suggestions and feedback. So it's a great example of how diversity can really accelerate the building of great software.
Bill Detwiler: I'd love to get your opinion on something I have been talking with leaders in the tech space for all of the last year, which is as a result of the COVID pandemic and many people working remotely, or many more people working remotely in the software development industry and in the tech industry when they can, that it's actually expanded opportunities, because it's forced leaders to consider people who maybe weren't in a certain geographic location. Not a lot of companies always did this.
Most of the companies that I talk to are global companies these days, have offices around the world, but there were still geographic barriers, or there were still psychological barriers that were preventing maybe as many people from being hired, or from participating as possible, but it seems that in the last year, there's been a little bit of a see change in that thinking. Have you seen that? Have you seen that in the open source community, that there is maybe more of a willingness now to accept input, especially on commercial endeavors, from a wider range of people?
Dries Buytaert: Yeah. I've definitely seen it. First of all, I think open source is kind of inherently open to the kind of input from people all around the world. From the early days, for example, Drupal had a lot of participation from people in different countries all around the world. Having said that, though, like have really seen a lot of open source projects, Drupal included, making real efforts to increase diversity and inclusion. Just being open source doesn't mean you're a diverse project. The data suggests otherwise, other than maybe geographic diversity. So at Drupal, we've done a lot of different things. It's a long list of small steps, let's say, all the way from changing our user forms, let's say, where we would ask for your gender. We would ask male, female, sort of coming to the realization that gender isn't binary and that people can self-identify differently, so adjusting our forms that way has been a step. That's more inviting to people that don't identify as male or female.
It's one element of diversity, to our conferences, where we purposefully give time to people from underrepresented groups to present. Give them time and space to show what they're doing, to even paying for their attendance often, inviting them to come to our conferences and giving them speaker slots. So these are just two examples, but we have a list of things that we're doing to bring awareness, and to really help improve our diversity and inclusion. If you look at the data, you see that it's working. Year after year, Drupal's diversity is kind of inching up slowly, and arguably too slow, but it is slowly but certainly getting better. It's been a consistent trend for a few years now, which is great, but we still have a lot more to do too.
Driving change from the top-down and bottom-up
Bill Detwiler: Yeah. I'm curious, as someone who is committed to this within your own organization, how much of it, in your experience, is driven top-down, or is driven bottom-up, or is it really a mix of the two, in that as the leader of the organization you set the tone, you say we're going to do this, you say that this is important to us and we're going to hold our leaders accountable for this, and we're going to hold our individual contributors accountable for this, that we're going to set real goals and metrics. That's kind of the top-down approach. And that provides, I think, some of the impetus for people to sort of... And the buy-in for people to do this.
But then there's also things that have to happen from the bottom up in that you do have to get those hiring managers. You do have to get the people, like you said, who are planning your events, to take what you're saying seriously and to then take that action, and to handle their unintentional biases, and give them tools to kind of overcome that, give them the freedom to change those forms, so that they're not in some way causing some kind of unintentional bias. What's been your experience with that? Because I'm hoping that other organizations, who as they're looking to do the same, either small, or large, or open source, or commercial, however it is, that they can learn from your experience.
SEE: Tips for building and advancing your leadership career (free PDF) (TechRepublic)
Dries Buytaert: Yeah. I think you said it well. So my opinion, it has to come from the top-down, but also from the bottom-up. So I think it's important for those in leadership roles to recognize the problem, because if you don't declare it as a problem, it's hard for others in the organization to feel motivated to help fix a problem. So I think it's good to say, "We're not good at this, and we need to do better," and then to set the tone and be an example of what we expect from others. I think it's very important for people in leadership roles to do that. At the same time, everybody needs to participate, because it's, as I mentioned, a lot of small actions every day, like biases, that we need to be aware of as individuals, and change our behaviors and make an effort to think differently, to behave differently. So I think leadership matters a lot, but at the end of the day, it's really the result of everybody collaborating and working towards better diversity is what will drive results.
Bill Detwiler: And that's what you're really after, because it's not enough to go through the motions. You really are after, and you've mentioned it several times, data. You're after the result. That's what's really important, isn't it?
Dries Buytaert: Right. For example, as a leader at Acquia, as I mentioned, we're over 1,000 people. As a leader, I don't interview every engineer. So what's important is that everybody that's involved with recruiting, not just the hiring managers but individual contributors that make hiring recommendations or decisions, that they understand their biases, as an example. So as a leader, I can't tell people who to hire and not hire on a day-to-day basis, because it would be too many. It just doesn't scale, right? So that's why I think everybody needs to be aware of their biases. Yeah. It's a team effort.
Going beyond the code: Making a great product
Bill Detwiler: Yeah. So let's talk about something else that has changed a lot in the last 20 years. And since we're talking open source, I think this is a great place to kind of finish on, is really talking about what I think a lot of people think of when they think open source, which is code. Code is the end-all, be-all. You started as an engineer, you started writing code, and so you think about open source in the context of code. I think as you put it, in the last 20 years, code really isn't everything when it comes to open source, is it?
Dries Buytaert: It's not. No. I think everything starts or stops with having a great product, or the lack thereof, if you will. So code is very important, right? I think for an open source project to be successful, it needs to work, it needs to be a great product, needs to solve a real problem, and a code is a big part of that. But having a successful project also means doing promotion and marketing, making sure there's funding for the project, that you can pay the bills, whether it's paying for servers, or paying for legal advice, whatever it is that you need to pay for, trademarks maybe.
So if you look at the successful open source projects, they're actually very well-rounded, meaning... And to have functions like those that you see in proprietary organizations, like you have an engineering team writing code, you have product teams thinking about the market, and go-to-market approaches, messaging, positioning. You have to deal with finances, because a large project like Drupal, we have a lot of bills every year, so we have to think about finances. We have a nonprofit that we're running as well, as an example, so there's the whole part of running a nonprofit, to HR, in a way, because even in open source projects you have to deal with harassment issues, or two people having a falling out, and what do you do? How do you try to mediate situations like that? So it really is all of the functions that you would expect to find in a traditional organization, that you have to establish as a larger and as a successful open source project.
It's something I wish I had came to realize much quicker, actually. Kind of learned it along the way when we had needs, and we have to like, "Oops. We need to really think about how do we solve these kinds of problems." So in a way, that's not a bad way to go about it, but knowing what I know today, I think I would start Drupal with more of these things in place from the get-go, because it would have avoided some more painful moments.
Bill Detwiler: It's the growing pains.
Dries Buytaert: Growing pains. Yeah.
Bill Detwiler: What about sort of focusing on the product, too, because I know that's something you and I have spoken about in the past, which is beyond the code, you really have to think about the user experience. You have to think about having a product that, as we've been talking about, meets the needs of as many people as possible, is accessible to as many people as possible, because sometimes I think folks who are very technical oriented, so that like to solve problems, that like gadgets, that like fiddling with things, that they can get caught up in that and forget about the sort of form. They can forget about the usability, because wow, I just really want to make sure that this opens faster. I really want to make sure this form works this way, or the code produces this result. We'll just throw it on a monochrome output and that's it...
And then on the other side, you get the people who are all focused on wow, this looks great, and this is a perfect feature, and I don't know how we're going to build it. So talk a little bit about maybe that and how things have changed in 20 years for Drupal around just usability, and maybe open source in general. It's kind of like open source, to me, has grown up a little bit, I guess. I remember the old days where it was like if you loved the command line, it was great. But if you wanted a GUI, not so much.
Dries Buytaert: That's right. Yeah. And by the way, I don't think either one is good or bad. But I think for me, when I started Drupal, I was a developer, as I mentioned, and I built Drupal for myself. I was just focused on the codes. I did not care about the user experience because I was the only user, and I knew how to manage the codes. I think for me, what I learned is that maybe that wasn't the right thing to do, because once I made Drupal open source, it attracted similar people, other developers, basically, and it created a little bit of tunnel vision, for lack of a better term, in that it was built by developers for developers. And for a long time, for years, we didn't really think about the user experience too much. But as Drupal grew, we saw that that was actually limiting, that that was holding Drupal back, so that slowly changed the culture. I know it's not always easy.
I remember having a lot of heated conversations about, "Why would we care about marketers, about less technical content creators?" And not everybody in the Drupal project was on board with that, and that's not necessarily a bad thing. I think it's fine that some open source code is just for developers. But I think in our case, it was important, because there was a long, multidecade long market trend, where content management systems were evolving from something that developers used to somebody that less technical people used, and that was incredibly empowering, the idea that everybody could self publish online as individuals, or the idea that you can enable marketing teams to build an online presence without having to involve developers, or having to involve them less. That was a very powerful idea, but that's more of a product management idea, if you will. That's trying to go after a market and an opportunity and create something bigger.
But anyway, I think what was hard for us for a while was changing that culture, changing that mindset of caring about people other than ourselves, and I say that sort of as a developer. But today we've made that transition, and we're very focused on user experience, and we've already talked about our focus on accessibility and all of these things, so I'm very happy and proud that we've made that transition, and we'll continue to work on that. But that was hard. So if I were to start from scratch, I would definitely start with thinking about all of the personas of my software. Developers still is an important persona, and user on Drupal is developers, but also non-developers, really designing something that all of the end users are persona in mind is one of those lessons learned.
Bill Detwiler: It's one of those things in hindsight that you look back on, and now it seems so-
Dries Buytaert: Obvious.
The next two decades of open source
Bill Detwiler: So obvious. That's right. So where do you think open source, where do you think you're going in the next two decades?
Dries Buytaert: Yeah. So I guess there's maybe two parts to my answer. First of all, I think Drupal will still be around, and I believe that because we're focused on innovation and change, but also because we're solving a forever problem, meaning 20 years ago when I started Drupal, people and organizations needed to manage content, and they needed to build a presence online and participate in the web. That was true 20 years ago, that is true today, and guess what? That will be true 20 years from now. People will still need to manage contents, people will still need to publish that content on the web, and do all the things that Drupal does. Never going to go away. So if Drupal keeps innovating, we'll be around.
I think for open source, if I step back, and I've kind of alluded to this, but I think today, open source has won in many ways, meaning I think very few people will fight me on the notion that open source can produce great software, stable, scalable, secure, which wasn't the case 20 years ago, as we discussed. But today, so many people use open source. So nobody doubts the open source development model. But I think what I would love to see us get better at is this notion of the business model of open source, the economic system might be a better term than business model, behind open source. And Drupal is a leader in that. We're doing a lot of interesting things. It's one of the reasons why we're so big.
But I really wish that in the next 10 to 20 years, we figure out the economic systems of open source to help us sustain open source projects, to help us scale open source projects, and to help us make sure that open source projects can be around for decades, or even 100 years. To do that, you need to have the right, I call it economic system, to fund it and things like that. I think that's a big unsolved problem. I don't think many, or any, open source projects do that really well today, and I believe if we can solve that, it removes the last barrier for open source to really, truly take over the world, so to speak. And then I think all software can become open source.
Dynamic Developer interviews and more
- uBreakiFix co-founder says new mobile repair service is about creating a great customer experience (TechRepublic)
- Quantum computing: Aliro wants to make quantum hardware more accessible for software developers and network engineers (Techrepublic)
- What software developers should know about 2021: Low-code, AI code testing, COVID-19's lasting effect and the skills needed to stay on top (TechRepublic)
- Amazon Alexa: How developers use AI to help Alexa understand what you mean and not what you say (TechRepublic)
- Programming languages and developer career resources (TechRepublic on Flipboard)