Open-source seems to be the uncontested standard for software startups. Yet investors are leery of investing in the OSS business models of yesteryear (such as Red Hat’s), and some of the biggest recent open-source successes — GitHub, GitLab or even Amazon Web Services — have grown more restrictive in the licensing of their stack. To paraphrase Tom Goodwin: Amazon, the largest open-source company in the world, open-sources none of its stack.
We caught up with Mårten Mickos, former CEO of MySQL, and current CEO of HackerOne, to talk about how Cloud & Containers changed open-source, whether a company like MySQL can be reproduced today, and why the most successful open-source companies don’t seem to open-source much software at all anymore.
Tim Anglade, Executive in Residence at Scale Venture Partners: So, this is lovely because I have seen the transbay terminal like build up from the ground level, but you guys are actually
Mårten Mickos, CEO at HackerOne: Yes, isn’t it fantastic?
Tim: Overseeing it. It looks insane. It looks even more insane than it looks from the ground. It’s just like, I, you know, I can barely imagine what the sea’s gonna look like with it.
Mårten: Well, every time I see it, I get this need to go and buy Legos and start building something myself because this looks exactly like Legos and Lego construction.
Tim: Right, and it seems so, like, intricate and like, the lanes and how it’s kinda merging. I mean, the view is kinda amazing. I think maybe people can just find that on Google maps or something or Google Earth, but that layout is just kinda crazy. It’s kind of insane. And, it must be kind of cliche but inspiring, right? You’re seeing all this activity and all this stuff building up as you’re building your company. Does that give you a lift at all?
Mårten: In the romanticized version, yes. But when you are in the office, your head’s down working, you’re on the phone, in your email, on your Hangouts.
Tim: Yeah.
Mårten: We never look out the window, other than when we have guests.
Tim: Well, if you do, maybe it is demoralizing because you’re moving ones and zeros, and people are moving like real blocks of concrete.
Mårten: No. It’s a good analogy. It’s a good analogy.
Tim: It works out.
Mårten: Yeah.
Tim: Okay, so we’ve met a couple of times before and we have had conversations about I guess building businesses for developers and around Open Source in particular. And, you know, you had the experience of doing MySQL, of doing Eucalyptus amongst many other things. You’re doing Hacker One now. You’re very, very well versed. I think the ideal person in my mind to talk about this. I feel like it’s an important topic to get into because it’s a central struggle. It’s one of the few things that I know about in startup where there’s this central struggle to make business and how Open Source kinda forces you to make this a certain way, but also hurts your ability to make business in other ways. So, it gets very, very complicated. But so, let’s start simple. Let’s start like, you know, do you think people could do MySQL today for example as an Open Source kinda plan business? Plays like right at MySQL. Do you feel like that is still something that’s still possible in this day and age or something that’s kinda largely gone?
Mårten: I think the world has changed in the way that most of the new software deployments are on the cloud and you do not deploy your own infrastructure anymore. You get it from the cloud provider. And, that was different when we built up MySQL. Everybody deployed their own thing. And they talked about downloading the LAMP stack.
Tim: Right.
Mårten: Now, if you ask a developer, they upload an image.
Tim: Right.
Mårten: And they create an instance. So they don’t even download. They don’t think about distros. So, I do think the world has changed completely. Open Source is still the superior method for developing software but you couldn’t build a new MySQL today.
Tim: It’s interesting, because it used to be, I know, I used to think of Open Source as about transparency and bug squashing and things like that but it turns out, as we move to that maybe it just had to do with delivery. It just had to do with the ability to download this and run it wherever you wanted in the easiest possible way. As opposed to some grander ideal about how your bugs are gonna be found or how many people were maintaining this or how transparent your software was, right? ‘Cause the more we go into those closed ecosystems like the cloud, you know, it seems like the less of a natural pressure there is from developers to use Open Source solution. That’s how you see the rise of stuff like Toleo, for example and Amazon itself, right? As a big kind of closestack that millions of developers build on. Do you feel like that’s part of that struggle?
Mårten: Well, I think first of all, that, every great phenomenon has many reasons for the greatness. So when you say what is the key thing about Open Source, you can list 20 different amazing things. And, one thing was, used to be delivery. So it was a smarter way to develop the goods and a smarter way to deliver the goods. Today, the delivery is different but, it still is the smarter way to develop the goods.
Tim: Right.
Mårten: You get better quality code, you get more people engaged. You get better innovation. You get a better integration. People can use it for whatever they need it for. So lots and lots of benefits with Open Source. But the actual delivery has changed, and therefore, the business models have changed.
Tim: Right, and so delivery, the forces were largely stuff like, a cloud and images and your know, like containers in general, and micro services and kind of what’s happening on that front. But the building of software, and let’s get that out of the way. We both like Open Source as a way of building software. That’s kind of very nice to have that community feeling and it’s a very personally satisfying way.
Mårten: I don’t do it because it’s nice I think do it because it’s superior.
Tim: Yes, okay.
Mårten: I’m happy that it is nice, but it is not an emotional decision. It’s a pragmatic decision.
Tim: Right.
Mårten: It’s a pragmatic decision for those who contribute and it’s a pragmatic decision for those who use it or build a business on it.
Tim: Right.
Mårten: Then, there are additional benefits, such as nice people and you’re having fun, but they are bonus benefits.
Tim: Right.
Mårten: Not the core reason.
Tim: So, let’s dive into that. I feel like a lot of people struggle with this. If it’s both a nice and a pragmatic and the best possible way to build software, why is it so difficult, comparatively speaking, to make money off of it? You know, if it helps you build the best software, why doesn’t it help you build the best business?
Mårten: Why would it be? Why would those two be connected? Many of the greatest things on this planet are difficult to build a business on. It’s great when people teach each other. It’s difficult to turn it into a business. It’s great when people take care of each other’s health medically. It can be difficult to create a good business around it. It’s great when people build Open Source software. It’s more of a challenge to figure out how to make money off that particular thing.
Tim: Yeah, and there’s no–
Mårten: But then you connect it. You connect it with something else. With the delivery or the packaging or the benefit of the service, and that’s your business.
Tim: Right. And, I am playing cool here, but, you know, I feel like that’s the thinking that a lot of people have coming into this business. You know, like technical founders come in and say you know, we’re building this software in the best possible way, an open way, we’re building the best possible software, money will happen, right? But no, you need to add that business component, right?
Mårten: Right.
Tim: Do you feel like Open Source makes it harder to add that business component or do you feel like it’s just as easy or hard as anything else?
Mårten: Open source makes different things harder and other things easier.
Tim: What are some of those things?
Mårten: Well, it makes it easier to reach out to people and get the following and get people to use it. That’s easier. It’s more difficult to go to someone and say I developed this Open Source software and I would like you to pay for it.
Tim: Right.
Mårten: Because then they say, no. it’s Open Source, I won’t pay you.
Tim: But the building advantages, they’re kind of, aren’t they negligible at this point? ‘Cause, you know, you have so much technical talent. You can hire for relatively cheap nowadays. You’re talking about the cost of building software which isn’t necessarily that high whereas the ability to derive revenue from it, you know, that seems to be more of like where companies are hurting.
Mårten: In a way, but when you build Open Source software, mostly you build something that will have huge impact and it will be deployed in the millions or tens of millions or billions. So at that point, development is not easy. It’s really hard. And you have to be the best, the fastest, the most lightweight, the easiest to use, the best replication, whatever it is. You have to be absolutely superior and at that point, creating the product is really, really difficult.
Tim: So its standardization right in a way, right? It’s kind of that ultimate goal that you can reach with Open Source, which is really, really hard to reach without having Open Source. Although it’s possible, right? I mean, you could say Docker reached that level of universality because it’s this gigantic Open Source community but Amazon became, you know, a big standard of cloud without really playing that Open Source card.
Mårten: Oh, they are playing the Open Source card completely.
Tim: Oh you mean like being compatible with that?
Mårten: RDS, RDS is built on Open Source, Aurora is built on Open Source. They run on Linux. They use the Xen Hypervisor. It’s a full Open source stack.
Tim: You’re right. I mean, I wouldn’t call it that. Maybe more of a co-opting of the stack? Would that be, you know, it’s kinda more it’s like we’re beginning to value that.
Mårten: That is a determination that we sort of have no rights to. Those who created Open Source released it under certain licenses, and if you use it, and operate it under those licenses, it’s fine.
Tim: Sure.
Mårten: Some people may try to call it co-opt and this and that, but it’s an irrelevant comment because, the intent of the creator was to allow all that action.
Tim: Right.
Mårten: So even if an AWS does not really share the improvements they are making, they are not violating any license, and they are actually furthering Open Source by making more of the world run on it. So I think we should fully accept it and say it’s a great example of how Open Source is revolutionizing the world.
Tim: Right, and you know, let’s maybe just call it hybrid. Right? They’re using this kind of Open Source acceptable footprint, you know, with something like RDS, but they are building their value-add
Mårten: I don’t think we need to call it hybrid. I don’t think we have to have a name for it.
Tim: Okay.
Mårten: I think every Open Source, you go down to the license and do what is allowed by it, and there are a million users and calling one more hybrid than the other is philosophically not useful.
Tim: Yeah, no that’s a very, very good point. I guess what I’m trying to highlight is that, I personally think there’s a lot of opportunities there in terms of products or companies that know to balance Open Source and Open Source-friendly and Open Source-compatible they are and how much a private value-add they are, whether it is in distribution, or in source code or in branding or something like that. And, I kinda feel like that’s kinda the key in 2016, of making money around those businesses, but I don’t know if you agree on kinda that thing or not.
Mårten: I do, I do. If people ask me, they say Marten, show me a great Open Source business. I would say AWS, RDS and the Aurora Services, great businesses that run on Open Source software. We can argue then about the purity of it, but to me, that’s irrelevant. Somebody has taken Open Source software and built a great business with lots of revenue and that’s fantastic.
Tim: You’re right.
Mårten: In the older world it was, and it still exists of course, but, in the older world, the dominant model was the one that Red Hat, Jaybus, MySQL stood for, and it was fantastic. But the world changes, and we must take the notion of openness to a new level and to the new world, and we mustn’t get stuck in some sort of sentimental view of how it once was.
Tim: Yeah.
Mårten: Because the world, you know, many things are bi-directional in the world but time is not, as far as I know.
Tim: Yeah, we’re not gonna bring it back.
Mårten: We are moving forward and whatever was in the past is only in the past.
Tim: Right, so maybe it’s not about that Open Source centricity or Open Source exclusivity any longer. It’s more about now that Open Source is in some ways a new normal, its all about the shades of gray and the balancing that you’re doing of how much Open Source you have in your business model and how you benefit from it and how you give back to it and how you build your business on the side and others.
Mårten: I think we have seen many other industries sort of build on the philosophies of Open Source. We share education today. We share genomes. There are a lot of sharing economy businesses, and not all of them are built on open principles but many are. And those entrepreneurs or governments or whoever drove it would never have had the courage or even idea to do so if they hadn’t seen what Linux and the LAMP Stack and others did in the Open Source space. So I think when we talk about Open Source, we must see that it’s broadening, and it’s being applied in politics, in governance, in all kinds of things. And, there are still more unexplored areas where we can be open, collaborative, where we invite people who don’t really know each other or like each other to collaborate on a common good.
Tim: Right.
Mårten: And that is very powerful and before us, for hundreds of years, nobody paid attention to it. That was the big mistake. Imagine! Imagine all these smart people who existed who couldn’t wrap their minds around collaboration but built these monolithic old-school, top-down controlled corporations that didn’t exploit the creativity and passion of human beings.
Tim: Yeah, and even without going into politics or anything so far-fetched, I mean, businesses like Hacker One, like that’s kind of built on that value system, right?
Mårten: Completely, yeah, yes.
Tim: Of being able to kind of distribute how much of your software is being built-in. And even if it’s not in that old-school MySQL, RedHat type of thing.
Mårten: Right. And I love Hacker One, because it has all the fundamentals of Open Source, of sharing, collaboration, delivering, distributing opportunity to everybody. Fortunately, it also has a business model.
Tim: Yes.
Mårten: Companies pay bounties, we give the bounty to a hacker. And the hacker is very happy.
Tim: Yeah. That makes a lot of sense. And I think that’s kinda the key thing, again, right? It’s like Open Source doesn’t deliver a business model for you, right? You have to always think about it.
Mårten: It’s a production method.
Tim: That makes a lot of sense. Thank you so much.