In episode 5 of Communications. Transformed., I spoke with Emil Ivov, founder and project lead at Jitsi. 8x8 recently acquired the Jitsi team and technology, so Emil is now also my colleague. Emil and I talked about the unique business model associated with open-source technology, how Jitsi grew out his lifelong interest in remote communications and why open-source isn’t just for idealists.

Building an Open-Source Business

Jitsi (the name comes from the Bulgarian word for ‘wires’) came out of Emil’s graduate work, building a softphone that would work on ipv6. Jitsi was a tool he developed to help evaluate the softphone he was building, and he started it as an open-source project. But as time went on, the open-source Jitsi project started to attract more attention while his Ph.D. work seemed unlikely to ever be used outside of academia. After finishing the program, he turned his attention to building a company and a community around Jitsi.

Like other open-source projects, as a business model, Jitsi was basically a consultancy—the technology was available for free and anyone could use the technology and make improvements. Companies who needed customization, however, would hire Ivov and his team to create exactly what they needed using Jitsi’s core technology.

Here are some of the key takeaways from our conversation about building an open-source-based business, and particularly an open-source communications solution.

  • As the project leader, Emil works essentially as a facilitator who creates the framework that allows for collaboration;
  • Without the open-source community, Jitsi would never have become the powerful video-conferencing tool that it is now;
  • Open source projects are geographically diverse and are part of the large phenomenon that innovation is spread out rather than concentrated at a few large companies;
  • In order for an open-source project to gain traction, there has to be something in it for everyone—including companies and individuals who use the technology, contributors who make the technology better, and consultants who help companies customize the open-source technology to get exactly what they need.

Emil talks about ‘pragmatic open source,’ or the understanding that just because something is an open-source project doesn’t mean everyone involved shouldn’t have their own, selfish reasons for being involved—that is, in fact, the only way the project will succeed, he says.

Episode 5 Podcast Highlights

Want to skip to the best parts? There was a lot of good info in this podcast, but here are some not-to-miss sections.

  • 3:06: Jitsi’s origin story, or how a side project became Emil’s primary business.
  • 13:09: Some of the most interesting ways Jitsi has been used.
  • 15:00: How open source facilitates collaboration and innovation around the globe.

Listen to Communications. Transformed. Episode 5:

To learn more about the technology behind Jitsi and how an open-source platform became a leader in video conferencing technology, listen and subscribe to the full podcast below, on Apple Podcasts, Google Music, Spotify or your favorite podcast player.

Links discussed in the podcast:

Read the edited podcast transcription below:

Emil: 00:00 Our product basically moved away from desktop and became this browser platform. And once we did that, that’s when we really exploded in popularity.

Speaker 1: 00:13 You’re listening to Communications. Transformed. A podcast from 8x8, where we interview the latest thought leaders and innovators, who share their insight about the future of enterprise communications. Let’s get to the show.

Randy: 00:28 Welcome everyone to the Communications. Transformed. podcast, I am your host, Randy Ksar, social media manager at 8x8. This week, 8x8 announced that we acquired the Jitsi open-source video communication’s technology, and it’s highly skilled team of open-source and video tech experts. The Jitsi technology further extends 8x8’s cloud technology platform and adds to the company’s video collaboration capabilities. Some of the most innovate webRTC, products, and companies use Jitsi to support millions of minutes of daily usage as part of their meetings, messaging and collaboration product ecosystems. The open-source community has played a critical role in advancing Jitsi’s projects by validating its use in a diverse set of environments, and complementing the core team’s development. As part of this acquisition, 8x8 is committed to continuing to support the growing developer community.

Randy: 01:20 Today on the podcast, I’m excited to welcome Emil Ivov, Jitsi founder and project lead. Welcome, Emil, and welcome to 8x8!

Emil: 01:29 Hey, thank you Randy. Very, very pleased to be here at 8x8 and very pleased to be here speaking to you.

Randy: 01:35 And so, for those that aren’t familiar with Jitsi, can you describe to us what it is?

Emil: 01:39 So, Jitsi is a suite of communication [inaudible 00:01:42] that let you build and deploy very advanced, very high-quality meeting solutions. So that goes over the way from what we’re using right here, just a very straight-forward video interface for two people to talk, and it extends into more complicated things, like being able to be connected to a telephony network. Being able to connect it to recorders, live streamers. Being able to add meeting notes, transcriptions, all sorts of things. So, that’s, I guess, the power of Jitsi is we already have all of these pieces, it’s already in there, and it’s likely the most advanced, the most feature-rich open-source video platform. In fact, even … not counting only in the open-source world, there are very few platforms that can compete with Jitsi in terms of features and quality.

Randy: 02:30 Very cool. So how did you guys get started? And was it you that started, or what was the kind of timeline from way back when?

Emil: 02:37 That’s a trip down memory lane and it always goes. I’ve always been fascinated by the ability to talk to someone remotely. As a kid I really enjoyed the game where you have the two cups connected with a cord and you get to talk to someone, and that’s sounded magical to me. My favorite toy was a pair of walkie-talkies that my parents brought to me, and it was just … I couldn’t imagine that my brother is 300 meters away from me and I could talk to him and he could hear me! We were using that all the time, it was amazing.

Emil: 03:06 So when I got into college and started work in parallel, I really wanted to find an opportunity to do something with Voice Over IP. It combined two of the things I was most interested in, programming and remote communication. And, with time, that came. Now when I started my Masters degree, I was working with the lead of the network research team in The University of Strasbourg, so I thought I want to do this. I want to build the soft phone that uses [inaudible 00:03:32].

Randy: 03:32 And what’d he think?

Emil: 03:32 And I would like to, yeah I would like to be my diploma, my thesis for my Masters degree. He said, “Oh, that’s great!” At that point they were working on IPv6, and it was all about trying all sorts of different applications and seeing whether they could work on IPv6. So he’d say, “Yeah, just make it support IPv6.” And that’s how Jitsu started. Eventually I started a Ph.D. with them and Jitsu was one of the ways that we were using to evaluate the work that I was doing, just to test the various optimizations were doing on top of IPv6. And then the open-source project actually accumulated traction. It felt really great. It was, in contrast, actually to my Ph.D. work, that’s a different story. Because every Ph.D. student knows that there’s always this doubt this thing that I’m doing, what is it worth? What is it good for? Because it’s always so advanced in a certain direction. So ahead of the market and the usage scenarios that, your always wondering what is this thing for? And, specifically in my case, I was just depressed by how I’d work on something, it would be very advanced, very complicated, I would spend days, and nights, and weekends working on it, and then I would publish an article and it would just go in a cupboard and we’d forget about it.

Emil: 04:44 In contrast, in parallel, Jitsi was just growing, and getting more usage, and I thought, “This is where I want to go. I don’t want to … ” Well, once I was done with my Ph.D., I thought I don’t want to pursue in academia, I’d like to go and develop this thing. So that’s when others joined me, Yana, Damian were among the first ones to join. We started a small company around Jitsi that was just doing the very traditional open-source business model of, “Hey, here’s the thing, it’s open-source, you can do whatever you want with it,” and then someone would come and say, “Well, it does 90% of what I want it to do, can I pay you to do the remaining 10%?” And that’s how we got a business going. That’s how it all started.

Randy: 05:21 And where did the name come from?

Emil: 05:23 Jitsi actually means “wires” in Bulgarian. Back then it wasn’t exactly what it was today. It was all packaged. A lot of the libraries were the same that we built, but the packaging was kind of like a Skype-like interface where you get to add your different accounts with [inaudible 00:05:39] servers, Facebook, MSN, back then was popular, and you had all these connections which we were seeing as different wires. You had wires to different networks and that’s why we called it Jitsi. And also because it’s just a cool word, it sounds good.

Randy: 05:54 It’s unique. And you could get the domain name, too, so there you go.

Emil: 05:56 Yeah, we had got jitsi.org, in any case. Jitsi com is still out there, I guess.

Randy: 06:01 Okay. Let’s take a quick break and learn more about 8x8.

Speaker 2: 06:05 8x8’s communications solutions help businesses transform their customer and employee experience. With one system of engagement for cloud voice, video, collaboration, and contact center, and one system of intelligence on one cloud communications platform, businesses can now communicate faster and smarter to exceed the speed of customer expectations. For additional information, visit www.8x8.com, or follow 8x8 on LinkedIn, Twitter, and Facebook.

Randy: 06:42 Now let’s get back to the show. So in terms of video conferencing, when did that get integrated to Jitsi? Has that been-

Emil: 06:49 That’s also a funny story, because it started from, from a product perspective, it started entirely the wrong way for all the wrong reasons, because it was a cool problem to solve technically. And as someone who is more product-literate today, I would never do such a thing, because the whole point is, we were just cramming the software with features. We weren’t really building a product for users. It made sense that way, we were building a product for companies to recognize themselves into, and say, “Hey, this has sort of the subset of features that we have, let’s go and take it and just pay them to brand it for us and add a few small things.” So it was just a constant, “What else can we think of to add into here?”

Emil: 07:27 And so we started by having audio conferencing. So you would basically, on your phone, on your desktop, you would call a bunch of people and you could merge them in a conference with no limit to how many people you could have. And that was an audio conference. And then we thought, “It would be great to add video on top of this. We should be able to do that.” Now obviously if you are hosting a video conference on a personal computer, that cannot be very resource-intensive, so we need to find a very lightweight way to handle this. So we cannot be doing video mixing. Back then, everyone was doing video mixing for conferences, you would just receive all of the videos from all participants and than you would unpack, then decode them, then create a composite image, transcode them, send different versions to everyone-

Randy: 08:08 Sorry, just pause you for a moment. The timeline on this? What year was this? You guys got started-

Emil: 08:13 I think that was about 2009? I guess 2008, 2009, and there were already a few companies that were taking an alternative approach, which is let’s just not deal with the video at the central point of the conference, let’s just go and treat it as packets, and the server is just going to be a router. And we thought, “This is exactly what we need. We just need to relay the packets from person A to person B and C, and we don’t deal with the video. Every person decodes their own video and lays it out as they want.” And that had advantages from a user interface as well, because you have … the end user agent has control of the individual streams, so they can select who they want to watch. And these advantages are available to this date, even in Jitsi Meet.

Emil: 08:54 Back then the only people doing that were [DDO 00:08:56] and Google Hangouts and one or two other experimental implementations. We implemented that, and it worked great, and it had a lot of success for our definition of success. Back then we were a lot smaller. But it kept coming back to the fact that, look you aren’t hosting a conference on a personal computer. And it’s very smart, it actually works with the resources of a personal computer, but this still remains a user machine. It can go down at any point, people have all sorts of crap installed on their computers. And no judgment, I have crap installed on my computer, as well, but it’s there and it can come in and take away precious resources. And your internet connection at home is not good enough to send a bunch of video streams to a bunch of people. We thought this doesn’t on a desktop machine. So let’s just take that part, put it in the cloud, and just control it from the Jitsi client.

Emil: 09:51 And we did that, and then at that point WebRTC came, and we thought, “Huh, so, okay, our client,” which was the main bread-and-butter at that point, “is great, but now we have this server piece, and it goes so well with this new technology that works in browsers. Let’s try doing an experiment. Let’s add the few missing pieces that Jitsi Video Bridge needs,” Jitsi Video Bridge is what we called this router that we put in the cloud, “Let’s add what’s missing there and then let’s have some sort of experiment in the browser to see how that works.” Back then we did that with Philipp Hancke, who is to this day a friend and probably one of the world’s leading experts on Web RTC. So we did this collaboration and it worked beautiful. It was just amazing.

Emil: 10:32 There was all the promise of Web RTC and plenty of stories there. To tell you one, for example, is when we started this experiment with Jitsi Video Bridge running conferences for browsers, we were coming right out of this long sprint improving the user interface in the Jitsi Bridge client to render conferences better, to layout the videos … better order in the screen. And that was very painful to do with Java Swing, which is what we were using on the desktop, extremely painful, extremely time-consuming. And it wasn’t working well, and we were hitting butts with Java Swing, and it was this mix of native components and Java code, it was atrocious. And then we go and we’re checking out this demo, that with a couple of JavaScript lines of code, [inaudible 00:11:18] have this effect where, as you were alone in your conference, and your seeing yourself, as soon as the second person joins, you have this 180 degree rotation of your video that transforms into the other person’s video. And that’s two lines of JavaScript code. And I was just there watching this and thinking, “We just spent weeks and weeks trying to just get something to look normal. And here with two lines of JavaScript code, it just looks amazing.”

Emil: 11:43 And it’s not even about how useful it is to have people rotate in real time, it probably isn’t that much, but it was just this ease of development, this speed of development that was promising. And we thought, “Look, we should probably concentrate more on the web conferences part.” And Jitsi still gets to live on because all of this code is running currently in our different components, in our connectors to [inaudible 00:12:05] networks, our Jitsi video bridge, but our product basically moved away from desktop and became this browser platform. And once we did that, that’s when we really exploded in popularity, and people came and some would just install the entire Jitsi package with the front end, the back end, and everything. We code front end Jitsi Meet, which is what we’re using right now, as well, and some would use that. Others would build their own systems that would just take the Jitsi Bridge and build their own front end. All sorts of different use cases.

Emil: 12:34 And then a few years later, Atlassian, who was back then investing in HipChat and making that a great product, they came to us and asked, “Would you like to work with us, or not?” And that’s how our first acquisition happened. And that’s it.

Randy: 12:48 And now you’re with us. Now you’re with 8x8. Cool.

Emil: 12:50 And now a few years later, we’re with 8x8 where I’m confident even more exciting things are going to happen.

Randy: 12:55 Yeah, the story continues. In terms of adoption, I think a lot of people would like to know, what are some of the use-cases and innovative ways that you’ve seen Jitsi being used?

Emil: 13:06 So, obviously it’s always about some sort of video communication.

Randy: 13:09 Right.

Emil: 13:09 But there are very interesting sorts of video communication. So there was a South American platform that was using this for communication with doctors. Connecting patients and doctors, and getting multiple participants to join in and communicate with the doctor.

Randy: 13:24 So [telemention 00:13:24].

Emil: 13:24 Yes, telemedicine is one. Then obviously we have people just using it in various sort of meeting applications, and those can become very exotic. For example, there’s this, I’m probably going to massacre that because I don’t really speak Spanish, Los [Spanish 00:13:39], or something like that, where, the three Kings, there was this Christmas application that a group in Spain implemented to get kids connected to the three Christmas Kings, and to have conversations with them. So that was pretty funny, too. And then we have educational institutions that are folks who are using this, we have some customer support scenarios, enterprise, all sorts of things.

Randy: 14:00 Cool. Any astronauts use it? Anything go from Earth to the space station?

Emil: 14:04 You know what, I haven’t heard of that, but that would really be great. There’s no reason why this wouldn’t work. So, who knows, maybe it even happened. But another field where Jitsi is particularly popular is in the domain of the privacy-minded people, and because it does offer you this unique, extremely easy way, with just a few lines of Debian commands, to get your own conferencing server deployed. Which means that you basically control the path of all the media, and you get the equivalent of a fully end-to-end protected conversation. So you would probably see Edward Snowden using Jitsi for a lot of his public appearances. That has happened on more than one occasion. In fact, he back up this Freedom of the Press Foundation, who actually have deployment scripts to facilitate journalists to talk to sources in various countries with difficult regimes. So that’s a good way for them to protect their privacy.

Randy: 15:00 Interesting. Let’s talk about open source. I mean, you talked about the community, and I mean your the facilitator to make all this happen. How has open source helped to make Jitsi what it is today?

Emil: 15:10 It cannot be overstated. There’s this book that came out about 10, 15 years ago that was called Innovation Happens Elsewhere. And the premise of the book was that you had these innovation centers in the past that were just huge. You had AT&T, Bell Labs, all of these, and today you have Google, and Facebook and 8x8, and a bunch of other places that group a lot of brains, a lot of brain power. And the premise is that no matter how many smart people you get in one of these places, there are always going to be more outside. And this is what open source lets you leverage. It is like … the way to think about open source is like this very easily, predefined way of collaborating.

Emil: 15:57 To give you a comparative example, the European Union is doing a really good job funding research project and collaborations between industry and universities. And setting one of those up is a multi-month, sometimes multi-year exercise. Where everyone gets together, get to define the project, gets to define the licensing, the terms, what happens if this succeeds and it doesn’t. And it’s an extremely complicated process. Open source, on the other hand, kind of preempts that process and says, “Look, we have this thing here, if your interested, these are the terms, these are the licenses, jump in and collaborate and contribute.” And that’s what happens.

Emil: 16:32 Now obviously it has to make sense for everyone. This is not a purely idealistic thing. In fact, I’m a big fan of pragmatic open source. It has to serve a purpose for everyone. The people who are maintaining it, the people who are paying full-time people to work on it, it has to be good for them. But that’s not incompatible with it being good for a bunch of other people, which is what we’re trying to achieve with Jitsi. So everyone has to be mindful of the other people’s requirements, and everyone has to know that they’re not the first and only user and use case that matter. As long as you get this right, open source becomes this great way of validating a bunch of things that you wouldn’t have necessarily thought of. A quick way of getting help, whether it’s in the form of comments, ideas, interesting scenarios and free work, if you really want to take it down to that, as well. And not to talk about, just the personal gratification of collaborating with other smart people out there that you wouldn’t have necessarily get the chance to do that otherwise.

Randy: 17:26 Yeah, I’ve dealt with open source over the years when I used to work at Motorola in the Android community, and so definitely I’ve seen something where it starts with sort of the baseline lowest common denominator, and it just evolves into something that you would never even think of, so, from an innovation standpoint, that’s definitely something that I’ve seen in the past. So, that’s really cool. So we definitely are … at 8x8 are definitely big open source proponents, too.

Randy: 17:49 Thanks for listening to part one of our interview with Emil from Jitsi. Listen to the next episode for part two of our conversation where we dive into how your business can incorporate open-source technologies, and finally learn a little more about Emil. To join the conversation online, join our LinkedIn group at bit.ly/8x8podcast, that’s B-I-T dot L-Y slash 8-X-8 podcast.

Speaker 1: 18:12 You’ve been listening to Communications. Transformed. To ensure that you never miss an episode, subscribe to the show in your favorite podcast player. Thank you so much for listening. Until next time.