Open source startups are not something you associate with Italy. Great art, music and architecture; fine food and wines; stylish design: maybe, but companies constructing masterpieces of code, no. And yet Funambol is precisely that: an increasingly successful startup writing free software for mobile phones and the backend systems that serve them.
The man behind it is Fabrizio Capobianco, who founded the first Italian Web company, Internet Graffiti, in 1994, and has had a rich and varied career spanning academia and business. He is also a well-know in the mobile world for his blog, Mobile Open Source. Here he talks about the tightrope that his company walks daily, the future of mobile open source, and why he is glad that the Honest Public Licence he drew up in 2006 has finally been superseded.
GM: When and how did you set up Funambol? Why are its headquarters in the US, even though you are Italian? Would things be any different if you were starting today, and how can we make Europe a better place for startups?
FC: Funambol was founded in Italy in 2002, following the creation of the Sync4j (synchronization for Java) open source project in 2001. In 2005, we received venture funding in the US and moved the headquarters to Silicon Valley, even though the vast majority of our R&D remains in Italy. The company name is derived from the Italian word 'funambolo', which means tightrope walker. We chose the name Funambol as we strongly believe that we must constantly balance between the needs of our open source community and our commercial customers.
We moved to Silicon Valley in 2005 but kept our R&D in Italy to get the best of both worlds: access to venture capital and world class talent for rapidly growing startups, combined with Italian design excellence and a passion for providing an innovative mobile service for the mass market. The mobile industry was further ahead in Europe than in the U.S. so it made sense to keep our R&D there. Our company has come a long way and we are seeing major benefits of this as our business really starts to expand. If we were starting over today, I would do it exactly the same way, for the same reasons.
As to making Europe a better place for startups, that is one of my goals. It takes a combination of venture capital, a culture of entrepreneurship and a critical mass of technical talent and education. We have some of these pieces in Europe but we need more of these.
GM: What problem is Funambol trying to solve? How does it work? Why is open source useful in this context?
FC: There are 3.3 billion people on earth with a mobile phone and there are two billion email accounts yet less than 2% of mobile phone users get mobile email. This is because until now, mobile email required smartphones, was too complex and was too expensive. At Funambol, we are bringing mobile email to the mass market. This requires that mobile email work on feature phones as well as smartphones, that it is very easy to use and that it is extremely affordable.
Open source is the only approach that makes this possible. Proprietary software approaches just cannot keep up with the rapid pace of change in the mobile industry. One billion new mobile phones are shipping every year. These phones have different operating systems, firmware, languages, etc., and operate differently depending on what mobile network they are used on. This results in a combinatorial explosion of devices that need to be supported and tested. Hiring a team of engineers in every major country and buying all the top phones would still only scratch the surface of the phones that need to be supported.
With our open source software that has been downloaded over 1.5 million times by 10,000 developers in 200 countries, they test our software on tons of phones and if it doesn't work, they tell us or they fix it, since they have access to the code. This is what we call the world's largest mobile QA team, it is our open source community, and it provides the broadest device compatibility, enabling Funambol to support many more phones than proprietary alternatives.
We have also made tremendous strides in usability. And we are pioneering a new ad-based mobile email deployment model that makes mobile email available at an ultra low price point which, for the first time, brings it into the range of mass market affordability.
GM: How might enterprises use Funambol? Who are some of your clients? What business model do you employ as an open source company?
FC: Enterprises are looking for a lower cost alternative to BlackBerries so they can deploy mobile email for more workers. They would also prefer one mobile email solution that can support all types of phones, including BlackBerries, iPhones and feature phones already used by their workers. Funambol enables people to access their mobile email as well as PIM data – contacts, calendars, tasks and notes – on their existing phones, for a fraction of the cost of proprietary alternatives. Our focus is on the mass market.
Our clients include leading mobile operators, service providers, online portals, device manufacturers and software vendors. In the past several weeks, we have signed up one of the top three portals in the world, two of the top five device manufacturers and a large U.S. mobile operator. An example of a customer is 1&1, the largest web hoster in the world with 71 million mailboxes. They recently went live with our open source-based push email & PIM sync service, for small and medium businesses, initially in Germany but with plans to expand to other parts of the world. Our customers tell us that they chose Funambol due to open source, there is just no substitute to being able to work with the code and control it. They also like the financial terms that we are able to provide.
In terms of our business model, we use a dual license approach. We have a free open source version of our software, and a commercial Carrier Edition version, which has the functionality to deploy Funambol for the mass market. Carrier Edition includes scalability, load balancing, fault tolerance, over the air (OTA) configuration of devices for mobile email and much more that our commercial customers need to succeed. It's a model that well serves both our open source community and our commercial customers – the funambolo is balancing quite well so far.
GM: How do you expect Funambol's products in particular, and the world of mobile open source in general, to evolve?
FC: In 2008, Funambol's software will be deployed to many more users around the world. As such, our software will be enhanced to support even more devices, to be even easier to use and to have more mobile capabilities. For example, we expect to provide important new releases of our software, that we develop in conjunction with our community, for iPhone, Android, BlackBerry and other devices in the first part of 2008. You will also see integration with mobile advertising as a way to make the mass deployment of our software highly affordable.
Mobile open source overall is gaining major momentum. This is due not only to our software but to Android, which is generating a lot of interest amongst developers. It is also the result of the trend of the opening up of mobile networks, which is partly being fuelled by Google and partly by the impending 700Mhz spectrum auction in the U.S. Mobile open source is starting to look like the internet did in the late 1990s, with an explosion of innovation and new offerings. It's exciting because we've been at the centre of this for five years and now others are joining us.
GM: At first glance, the mobile phone market seems ripe for commoditisation, with Linux as the free operating system underpinning everything. Why hasn't that happened, and will it ever happen? What effect will Apple's iPhone and Google's Android platform have on the situation?
FC: As mentioned, the mobile phone market is huge. Although you may think of a mobile phone as being primarily used for voice calls, texting or maybe a PDA, it's becoming the primary computing and communication device for many people. As a result, the market for mobile phones is extremely diverse, with many different types of users around the world. There will continue to be many choices for different types of mobile phones, just as there are many different types of cars 100 years after they were invented.
This choice requires there to be different mobile operating systems (OSs) to meet the needs of different users. That said, mobile Linux is definitely gaining momentum and market share at the expense of other mobile OSs. The primary issue today for mobile Linux is fragmentation; there are several competing alternatives, which will be winnowed down by market forces over the next few years.
We do expect that iPhone and Android will have a major impact on the adoption of mobile Linux, as both phones are based on it. Apple just announced that they have sold 4 million iPhones in its first six months and we see this picking up steam like the iPod did. We expect them to gain significant market share at the expense of other device manufacturers over the next few years. Apple has a history of growing the iPod from single numbers of millions of units to tens of millions of units, and although there are different dynamics in the mobile industry, they look like they will achieve similar growth with the iPhone.
As to Android, it also has the potential to greatly shake up the industry. It could allow Google to provide low-cost or even free mobile services to the masses that are supported by mobile advertising. This could completely change the operating model for the wireless industry. The first production Android-based handsets are due in the second half of 2008 and it may take another rev or two of their software to make it right but they have deep pockets and are committed, plus they have the open source movement on their side. Together, iPhone and Android will make a substantial impact on mobile Linux adoption.
Funambol, by the way, already supports both of these platforms and complements them by providing an open source server for syncing all types of data and content, not just to these phones but to 1.5 billion feature and smart phones. So the more successful iPhone and Android are, the more current mobile operators and service providers will need a solution like Funambol to provide value added mobile services.
GM: A little while back you were worried about the effect of software as a service on open source: does the new GNU Affero GPL address those concerns? What about the future: do you think that the rise of software as a service will be a problem for the free software model, which is based around the idea of distributing code, not just using it over the Internet?
FC: Funambol has always been about making sure that the spirit of open source is followed. In our view, it is fine if someone wants to use open source software for their own internal purposes, they can do what they want with the code. But if they are offering it to the public, they should give something back to the community that helped created the open source, whether it is code or money.
This is why we created the Honest Public Licence (HPL) in 2007, which was essentially the same license as GPL2 except that we fixed a “bug”: if you "distribute" a service over a network (such as the Internet), it is conceptually the same as physically distributing software (such as via a CD), hence you need to contribute any source code changes back to the community or buy a commercial licence that obviates this need. This “bug” is known as the ASP loophole and we led the charge creating a licence, HPL, which closed the ASP loophole. It was edgy to lead this initiative, as there are many large companies that greatly benefit from open source software, that are making a lot of money from internet-based services, that are not contributing their code to the community as they see it as their proprietary IP.
We were very happy to see that the AGPLv3 licence, which was ratified late last year, addresses this issue. We adopted it for our software the day it became available. It is the same as GPLv3 except that it adds an almost identical clause as described above. We are in the process of submitting AGPLv3 for OSI approval to have it made an industry accepted open source compatible licence and we expect this will happen in 2008.
We definitely see Software as a Service (SaaS) increasing over time as it provides very compelling benefits to enterprises. Regarding its impact on open source, this is exactly why we advocate the use of AGPLv3, it will allow open source to thrive in a SaaS environment.
The definitive collection of interviews with the key players in the open source and free software sector.