A Master of Open Source

An interview with Frank Karlitschek, Founder and CEO of NextCloud

Colby: I'm Colby Gilbert, creator of The Opened Hardware Podcast. In this episode, we interview Frank Karlitschek, founder and CEO of Nextcloud. Frank has long been a key figure in the development of the open source community, including his work with KDE. And even as an advisor to the UN. Thanks for listening and enjoy the episode

It's hard to even pick where to start. I don't know how you find time for all this stuff, but to start just kind of your background and learning in school. And I always like to know what open source looked like when you first started 

Frank: Since I was really a kid, since I was seven years old. My uncle gave me a programmable pocket calculator, which really got me started with writing software. Writing software was a big thing because it had a one line display. So it's really hard to write complicated software if you can only see one line at a time. But that really got me started, and since then I really got into computers and really got into software and computers overall. Then after school I decided to study computer science, did this at a university.

And after that, I started in industry as an admin slash developer slash manager of development teams and all kinds of things. In parallel, I really got into open source in the middle of the nineties. This was really fascinating because until then, I only knew computers and software, where everything is done by companies. But then a friend of mine basically showed me the very first version of KDE of the KDE desktop. And I found it super fascinating from a software perspective, but also when I learned that this is actually not done by a company, but this is just done by hundreds or thousands of volunteers all over the internet.

Just volunteers who have never seen each other in person. Collaborating through the internet and building a software, which is super cool. And I just thought that this is the most amazing thing. Again, not only the technology, but also from a social aspect and society and how to work together.

Super fascinating, because it's just open, those communities. You don't even have the classic structures. You have a lot of companies, like with management and product management and sales and marketing and team leads and stuff, that don't really exist because everybody's a volunteer. So everybody does what they want to.

I thought it was really fascinating. And I got into the KDE community, was active for quite some time, and did all kinds of things there. Little bit of coding, little bit of artwork, little bit of event management. And I was a board member of the KDE foundation for a while, and was really deep into all kinds of open source topics. 

Colby: You did a ton of projects and you mentioned, even in the nineties, there were hundreds or thousands of developers involved in the project worldwide.

Frank: It's really fascinating because a lot of things changed over time. There was no Github.

Colby: I can't even imagine how you collaborated on projects. 

Frank: Yeah, exactly. The whole workflow was so different. There, there was a time where patches were sent around by email and everything in the organization happened by mailing lists. Mailing lists are not that big anymore nowadays. And there was nothing on the Github level, like a real collaboration platform didn't really exist.

It didn't have the sophisticated branching or reviewing. All this complicated or sophisticated collaboration, workflows and tools we have nowadays didn't really exist. It was mostly like one person or a few people had commit rights and they were just committing all the stuff. The reviews happened by mail or chat mostly. But it worked, it really worked. 

Colby: How many people were on like the actual KDE team to manage all of these, like thousands of people and requests and everything. 

Frank: I don't really remember. And, of course there were lots and lots of mailing lists. There were some core mailing lists which were super active and hard to follow. But yeah, there were hundreds of mailing lists, hundreds of ISC channels, quite complicated. 

Colby: And off of KDE, I saw that you had a few different projects that were spinoffs or directly related to KDE. Were those all like right under KDE, did you branch out and do those projects individually or with a team? Like the KDE open desktop.org, which is the one I was really interested in, those projects. 

Frank: Good question. For KDE and these open projects, I don't really have this hard boundary where something is in the project or out of the project. It was the ideology behind it. 

So the whole website thing I started was KDE look.org. This was the first one I thought of, because the development process that I described with sending patches by mail and ISC channels. This was really optimized for development of code. But if you really want to do something else say, create a cool wallpaper or cool sound theme or I don't know, something else more visual then I thought that there's a need for an easier way to collaborate and contribute something to a project or to the world. In the end I built it like in 2000 and I think it was launched in 2001. KDE look.org, which was I don't know, nowadays we call it a social network. It was just a platform where people can come together. You can upload something like a wallpaper or an icon or something, and then other people can comment on it, can read it, can like it, download it. Of course, groups and messaging and all kinds of things. It became a really big platform where people collaborated around the more visual side of open source. 

And then the GNOME project, which was always sort of the sister project, as KDE had similar goals, but existed in parallel. There was also a need for a platform like that so I launched GNOME-Look.org for them.

Colby: Now with GitHub and all the forks, it's hard to differentiate project from project really specifically. But for those projects, especially KDE look with the social networking aspect and everything, that's like a different animal. Were you developing these tools directly or were you more, at that point, in a managing role of those teams?

Frank: No. This is a once a one person operation. I did everything alone. I wrote the core software, hosting the server, which was not that easy because the traffic was huge, also paid for the server on my own, and had two banners on the website that paid for the server. This was a bit of my contribution to open source. And I was running that for a few years.

And then you mentioned open desktop, which was basically the umbrella. Because over the years, more and more projects came to me and said, Hey, I want to have a website like that. And I launched all these websites and it became a bit confusing and considered everything under the open desktop org umbrella, basically. 

Colby: This is a bit of a side question, but I like to ask how you learned all this, especially for you, because you had your hands in so many different fields, all these different languages and infrastructures. Now, open source tools, to learn them, there's YouTube videos and there's a lot of resources now, but I'm guessing when they first were made, the documentation was not as good as it is now. 

Frank: Oh, no, I feel really old, I have to say. It was possible to learn technology before YouTube. I mean I bought books. Books about how to write C, and how to write a sampler and how, I don't know, how TCPRK worked, and had all these books. That's how it worked. 

But of course, I mean the internet still existed. And there were website's, forums, where you could ask questions and exchange ideas and everything. This existed, but no YouTube and no stack overflow and no github. 

Colby: After those projects, did you move into roles like CEO and founder and roles like that, did that happen right after KDE? Or did you step into a couple other roles before?

Frank: This was a process. At this point in time, I basically had a split personality, because I was doing all this IT work in my day job where I worked for different IT companies and I was mainly managing engineering teams, management, customer relations. But in my other life, I was just an open source fan and did all that stuff for fun. Basically the KDE stuff was always a hobby for me. And then at some point I just thought, is there like an opportunity to bring this together somehow, to make a living out of open source somehow. This was when I founded my first company, which was like half successful I would say. But then a side project of that was Owncloud basically, which is an open source Dropbox, that's what I called it in the beginning, that you can host yourself.

It was also something I wrote like a hundred percent on my own to version 1.0, which was not that good at the time. I think none of my code still exists, which is good. But I did this and talked about it, built a community around it. And then there was also an opportunity to build a company around it. This is when I became like a real founder and CTO and we had some venture capital investment and became like a real big business. 

Colby: I'm glad you mentioned that, because for investments and stuff, it's often hard for investors to grasp how you can make money with open source tools because they're open source. So how did you explain it to them? Or pitch your idea to them? 

Frank: So for me, it's quite obvious. Because making money means that you're selling something, and of course you need to sell something, but you don't necessarily need to sell software licenses. There are lots of other things you can sell. Like you can like your expertise, support, for example other things. And there are good examples for that, red hat. At the time, there were a lot of successful companies who didn't sell software licenses, they didn't have to, the software is free, but there are other things they can sell, like helping companies to run it on a big scale, security patches, consulting workshops, all kinds of things.

Colby: And was that it for Owncloud? Was it a software package or were you offering something else? 

Frank: No. The business model was, the Owncloud times, it was actually a bit unclear. Because this was also one of the problems of Owncloud, that a lot of people had a lot of different opinions. What to sell and how this all should work, and this was the reason why, at the end, it was not successful. 

Owncloud was many things. Owncloud sold the support we just discussed, but Owncloud also sold an enterprise version, which was proprietary. But there was a community version, which was open source.

 But if you want to have all the features, you need the enterprise version and this is not open source and this costs money. And I always disliked that, because then open source is just used as a marketing tool. 

There were also experiments with dual licensing and all kinds of, I would say, not so good business models. This was also, you're right, this was partly influenced by the investors who didn't really understand it. And I also made some mistakes with picking my co-founders, to be honest, and at the end, this was also the reason why Owncloud was not successful. And then of course, the core team, including me, decided to reboot everything as Nextcloud and I think we made some better decisions about the business model and strategies and community and many other things. And I'm really happy that Nextcloud is now fully working. So we are profitable. We’re growing organically. We don't have any external investors. 

Colby: You don't have any external investments. Wow. 

Frank: No. This is all self-funded.

Colby: Off of that and managing and becoming CEO and all these different kinds of executive roles. I mean, you got the chance to do it early on, those roles, but for a lot of really technical engineers and technical founders, it's difficult to be a software developer and then the next day after you get money, you're in a leadership role, was that difficult? To go from working on technical projects to managing, or was it just a natural transition?

Frank: No, no. This is difficult. As you said, this is a process. For me, this really happened over 20 years. This really happened over a long time. It was really in the nineties, after my studies at my first job where I think I was just a developer for a few months and relatively quickly, I became the deputy head of department and the later head of department. So it's interesting how everything's developed and done, how it works technically, but then also explain it to customers and try to convince them that it's something useful for them. 

Colby: That's why that's usually why there's two founders.

Frank: And then also working with people and then this leads to management. I always thought, combining those two different things is interesting. Of course, in the nineties and the beginning, I made, obviously, so many mistakes and so it's a challenge, but then over years I just got better and better. And now, I basically don't write any code anymore at all. I'm only working with customers and team members.

Colby: So are you involved in a lot of the real business stuff and kind of seeking investment and everything, because that's also another major you can get in university too. Did you pick up that stuff as you went? 

Frank: This again was a process. There were some times, also at the beginning when I started with Owncloud, where I thought that this needs some magical expertise that I don't have, because again, I don't have an MBA or something.

And I thought I cannot do that. I leave this to other people which then didn't work out. And nowadays, I came to the conclusion that first of all, it's not rocket science. You can really learn it. You can get into it. Now the insight I have, is that it all belongs together somehow, you cannot have one person who thinks about business and the other talks about technology and the third about marketing and the fourth about community. It's all disconnected, which just does not work. This all needs to come together. What you sell is also what you build and what the community motivates. It's also something that you can use, and you have to give back to the community what they want, and it all needs to fit together somehow .You cannot really have one business guy and one technology guy and one, I think this is a mistake. It needs to have one strategy. 

Colby: Exactly. It's really hard to sell something if you don't understand how it works either. Off of picking up those skills, I wanted to talk about your speaking engagements and going to all these conferences, because looking through these conferences and looking at the talks, it seemed like you mostly did the high tech open source fields, like the Silicon conferences and everything like that. How did you first get involved with these, because you have a pretty big range between conferences and speaking at schools and with different audiences. 

Frank: It was a process. I really didn't like it at all 20 years ago, and then got an opportunity to give my first big talk. It was to the KDE community, I gave a keynote at the KDE academy conference. I just did it because I wanted to pitch my ideas and tell the story I wanted to tell. It was not about talking, this was not the purpose on its own. I wanted to communicate something and this is why I gave this talk.

And then another one, and then it somehow grew. For me, it's a part of communicating with people, exchanging ideas. That's what I find interesting. I don't think I'm someone who really likes to give talks. For me, talking is not the point. For me, the content is the point not just giving a talk. It's because I want to share it with people. I want to hear what they're thinking about the ideas. So the talking itself, I still hate, but I love to communicate with the community.

Colby: And off the content, do you stick with the same rough topic or area? Because you go to a lot of different conferences that are in various different fields. Do you kind of stick with the same talking point or not talking points, or the same areas of discussion, or are you having a different talk every time. 

Frank: For me, it's again, I'm not giving talks to give talks. For me, I want to talk about something that I find interesting or that we are working on, to communicate or share. So I don't have that many different topics basically. So for me, it's always about the importance of open source, privacy security, how software is written, what it can do.

 I'm not able to give talks about something that doesn't interest me. I really need to be passionate about the topic and then I give a talk. Also, always speak freely, I don't really write down the sentences I'm going to say. I really always like to just improvise, and this only works because I care about the topics.

Colby: That's much better than just listening to a lecture. And your talks and what you're focused on now, growing open source and everything. It's getting recognized and it's cool to see more and more larger companies actually be able to, and be willing to start using these tools. But obviously there's concerns about security and everything.

Because people just naturally think that open source means no rules and there's no maintenance or anything, privacy with the open source. How do you go about that? Because that's something where it's hard to find good resources and people looking into it. It seems like people are mostly focused on developing actual tools and not worried about privacy and security and everything. 

Frank: That's a good question. A lot of people have stereotypes about open source who think that it is just some nerdy things, which are not really usable for normal people. Which is of course obviously wrong. And then I often like to point out examples; hey, do you use Android? Yes. It's open source. Okay. Do you have a Linux server somewhere? Yeah. That's open source. Do you use any modern programming language, then they're all open source.

There are lots and lots of examples that you can give. And then all these open source tools can also be professional and useful. That's one stereotype. And you also mentioned privacy and security. This is something people need to be a bit more conscious about. What they're doing, how everything works, what it means, where it comes from, and so on. If you just go to some website, you use it, don't even think what's happening with the data, who has your password and so on. You need to think about it and know the transparency, like what code is running there, who has access to it, where is the data stored, and so on?

Then open source of course, very important because since it's open source, you can actually check what's happening. Is the encryption of my password good or bad? Or is there a back door somewhere in the code or something you can check with open source?

Colby: And which fields do you see growing rapidly? Because with Linux, it was operating systems and now there's a lot of open source operating systems and the chip design open source community is still relatively small comparatively, but it's definitely growing, especially with SkyWater and everything. Are you seeing other fields that weren't open source start to lean into open source and explore some of the tools or is it mostly still what it's been the last dozen years?

Frank: That's a very good question. Open source seems to grow from the bottom up. That's the kernel right? Linux is everywhere. Most of the devices we have use Linux; IOT devices, phones and connected coffee machines, cars, or planes. Linux is everywhere. And then of course nowadays you also have a lot of middleware, like services and infrastructure and libraries and all the things on top of it, the databases and storage systems and networking and application servers and all that stuff.

It's also like more and more open source nowadays. There's still some framework, I guess if you're into apple development, for example, then a lot of the APIs and frameworks are not open source, but for the rest of the world, most of the middleware stuff is open source nowadays. What's mostly proprietary still, is the application layer on top of it. A lot of companies think that this is their secret sauce they need to protect. I think this should change at some point, I mean this is also what we do with Nextcloud. Nextcloud is user-facing. Nextcloud is really on top of the stack. It is what people interact with. That's the area which is mostly proprietary and we based Nextcloud to do a full open source alternative. 

Colby: That would be awesome. Because that's one of the last areas that is still almost completely proprietary, and mentioning the UI stuff and everything, and your role. I even saw you were an advisor to the UN, which was something that I can't believe you're doing too. But making tools like that available to everyone, that's something I'm interested in too. Being able to use these for education, because in developing countries, I talked to a couple developers in Pakistan and India, they can't get the cadence and synopsis licenses for chip design. So they have to use open source tools. And now that they're at the performance they're at, it's possible to use them for education and everything. And for the UN and everything. What's your role with that? Are you focusing more on expanding open source or educating people on it.

Frank: The UN role I have, the advisory stuff I'm doing, that's not a big thing. The UN is developing some best practices, some policies, and some ideas on how to handle open source licenses in the right way. I'm just helping a little bit to build up these frameworks, how to do that.

Because as you said, open source is super important. For example, for developing countries. From the UN perspective, it's very important that if some money goes into a developing country, that it goes there in a way that it helps everybody. It doesn't go into the hands of a few rich people, and then most of the people get nothing basically. 

And this is of course where opensource can come in, because if you help them with technology, with software, and some funding to write software, or even funding a school that people learn how to write software. If all the software is open source, then it can be used by everybody and everybody can contribute and learn from it. And you can teach it in the next school because you can actually study how it works. This is why open source is so important. If you give some Oracle licenses to developing countries, this doesn't really help society, right? And this is where I'm helping there a little bit, but this is just a tiny hobby of mine. That's not me. 

Colby: Spinning off the future of open source and the future of your career. Because now, comparing it to your earlier career where you were doing all these crazy projects individually, you're a leader of developers and companies.

Do you think you're gonna try to pursue Nextcloud and speaking more? Or do you think you're gonna still keep your hands in all these different open source projects and everything? 

Frank: Of course. One of my goals in life is to have a positive influence in the world. I want to do something good. And I believe that giving a collaboration tool, like Nextcloud, to the world, which everybody can use. As you said, developing countries can use and study it. And without money, just put it on a raspberry pi and run it. This has some value. And if you care about privacy and security, I don't want to have surveillance then open source is very good. 

Without something like Nextcloud, everybody would use Microsoft or Google or Facebook or Dropbox. Then in five years, all the information or communication of the world is stored by five companies. And this would be weird. For me, this is the vision, to try to prevent this, to make sure that there is a free alternative to these big tech companies. So that's what I want to achieve. Those are my goals in life and all the other things I'm doing, like being CEO of Nextcloud or giving talks and other things, this is just to achieve that.

Colby: I would say that you've already achieved it. If I were you I'd be retired by now. But for Nextcloud, I forgot to ask this earlier, but currently, who are the customers and who's starting to use Nextcloud and become interested in it because it's always cool to see companies, and major companies, start to become involved with open source.

Frank: So there's actually very diverse, different kinds of users. First of all, Nextcloud has millions of uses all over the world, right? Because it's open source, everybody can use it. You just download it and just use it right. You don't need to register, create an account, or pay Nextcloud. That just isn't open source. So I don't even know how many users are out there, but from rough calculations, most likely many millions. But they're mostly home users or small users.

The big organizations, they probably come to us as a company because they want to have support and consulting, all the stuff companies want. We have some bigger customers. So it's a lot of governments, we have the German government, we have the French government, Swedish government now European commission lots of governments in South America, Asia. Basically organizations really want to keep their data secure. 

Colby: Getting recognized by governments is the top of the line for open source. Because usually they're the last ones who want to use open source tools. 

Frank: That's actually what I thought too at the beginning, I thought that clearly, we will sell a lot to companies and private businesses and government comes last, but actually governments really like what we are doing. So they're quite active.

Another big area is education. We have hundreds of universities all over the world using Nextcloud because they want to provide a collaboration environment to their students. And they often have their own infrastructure already. And for them, throwing Nextcloud on the server or cluster of servers and providing it to the students is relatively easy. We have a lot of companies like classic companies, enterprises, like I know Siemens or Infinian you might know as a chip guy. 

And several others and then also service providers, which is also interesting. So the German telecom companies are OVH in France or American mobile in Mexico, and several others are just cloud service providers who want to provide a service that competes with Google and Microsoft. And then rebrand it, usually under their own name and put it on the infrastructure and then they have something to sell to their customers.

Colby: This was really awesome to talk to you. I appreciate it.