Sunday, June 25, 2006

Open Source for the masses

The goal of Funambol is to create a platform that will bring mobile applications to the masses, starting with push email. Obviously, via open source.
Bringing open source to the masses requires a bit of explanation. If you work in this space, I know you tried to tell your mom what you do for a living and failed. She does not understand what open source is. In my experience, the problem started with her not understanding with "source" means.
Some time ago, I have been asked to hold a seminar around mobile open source and dual licensing to college students. The crowd was a mix of people with engineering and economic backgrounds. The first group, techies enough. The second group, a bit more techies than my mom (with all due respect) but still not enough.
The key is that you have to explain "source" first: non techies do not understand the concept of source and binaries, let alone compiling. We grew up with that. It looks simple but it is not an easy concept to grasp for the masses.
Hence, here is how I explained open source to them:
- how does Coke produce its final product (the beverage you like)? They start from a very secret recipe (the "source"). Based on that, they produce your Coke. The recipe is not available to anybody, it is totally proprietary. Nobody can improve it, but Coke. That's what Microsoft does with Excel.
- if you want to make Tiramisu, the recipe is public. It is not proprietary. The "source" is open, available out there on the Internet. Anybody can improve it and make a better Tiramisu than the person who first invented it (which I trust is not my wife, though her Tiramisu is great). Many today share their modified recipes on the Internet. Most do not and keep their changes secret (proprietary) to impress new boyfriends. Wouldn't be great if anybody who improved the Tiramisu recipe would be forced to share it on the Internet?
- THAT is what is behind Open Source. The recipe is public and open and available on the Internet. Everyone that changes it must share it with the world. The concept is called copyleft, which is something different than copyright (yep, engineers might love compiling source code into binary code, but are also funny people). Copyleft is the basis of the GPL, the most used open source license. You change something in the source, or bundle the source to distribute it with your code, and you have to open source everything. That is what makes open source software incredibly better than proprietary software.
- if you do not like the idea of open sourcing your code (you want to create the uber-Tiramisu) and behind the project there is a company with a dual licensing model, then you can license the code (in exchange for cash) and get rid of the copyleft requirement. That's how Funambol makes money and puts it back into the community to improve the product even more.
Lately, I found out that this explanations brings my point home. My mom knows what I do (and still calls me every time there is a problem with her computer). I am not sure if it will work for you as well, but if you try, it works and you are grateful, I accept Tiramisu as a gift.

Friday, June 16, 2006

Open Source and the Mobile Market (InfoWorld post)

This is a recent post that appeared on the InfoWorld blog "Open Sources by Dave Rosenberg and Matt Asay".
--- cut here --- cut here --- cut here --- cut here ---
We recently had the pleasure of hearing Dave rant about his mobile phone and BlackBerry death-spiral. Fortunately, he won't have to start rubbing sticks together to send out smoke signals any time soon - I promise.

RIM is starting to target the consumer space by inking deals with operators. The latest: Cingular in the U.S. In theory, this should make email easily integrated with cell phones... and RIM does have a pretty good email solution. Problem is, RIM's solution -- and their business model -- has depended on their proprietary hardware. They are stretching a device built just for the enterprise to adapt it to consumers. So far, only lowering the cost... That is not the way to do it. Consumers need a different type of device. Do you believe my father will ever use a QWERTY phone? I do not think so. However, he uses SMS today. And he receives emails on his desktop. One life, two separate worlds (mobile and web). I does not have to be that way. Good news: it is already changing.

How are RIM going to make their business work with all the phones from dozens of vendors in the consumer market? How do they extend functionality and interoperability to address this myriad of devices with proprietary software? They're smart enough to know that email is the killer app for mobile... reaching the mass market handsets (like your Razr) with email (or SmartSMS ;-) has nothing to do with being smart and everything to do with open source and open standards.

Open source software can do more for the mobile market than any other market it's disrupted to date. Why? Because there are 750 million phones in the world that are compatible with an open standard for PIM and, soon, mobile email (it is called SyncML). And because the only way to test all those phones (with the variations of software builds and network operators) is to harness the power of an open source community. The lousy customer service you were getting from the operators? Open source communities provide better support than that because of natural incentives for cooperation -- look how fast Linux fixes happen and the Linux device drivers history.

The trick is making sure operators "get" the idea of open source software as a real business value both on the cost and revenue side. It's not a religion or philosophy. We have some work to do but I know Dave can hang in there. Hold on -- we're nearly there. With open source and open standard your mobile life will change (and it will be finally integrated with your web life).

Sunday, June 11, 2006

Is SyncML slow? Who cares ;-)

Last week another person asked me: "is SyncML slow?". This is a question I get often, due to a campaign launched by some that claim to have a proprietary protocol 20% faster than SyncML.
Let me start answering the question with a couple of questions:
1. Is HTML the best language to design pages? I do not think so... It is actually pretty bad and it was even worst at the beginning. Though, it is easy to read, and learn. And it is a standard. It created the Internet phenomenon that changed our lives. AJAX has been invented to go around it and kick off Web 2.0 (not sure what it is, but it appears to be a cool thing).
2. Is HTTP the best protocol to transmit data? I do not think so... It is actually fairly limited, not fast but easy to read and learn. And it is a standard. Millions of users use it every day.
Now, what about SyncML? Well, it is based on XML. It has some overhead. That makes it easy to read and learn, exactly as HTML and HTTP. Is it a standard? You bet. It is installed today on some 700 millions devices.
Just because of it, I am starting to feel better about that 20% difference. I would go with an open standard over a proprietary standard any day. I would bet Stefano could write a protocol 20% faster than HTTP overnight. Unfortunately, it would not be likely that 700M people would use it. So we stick with the standard as we did in 1994, when we decided to go with HTML and HTTP.
One more thing about SyncML and, in particular, version 1.2 with push email: this is not browsing. You are not waiting for a page to download. You hear beep when your email is already on your device. The 20% difference means you will hear beep a few milliseconds before me. Big deal. SyncML might be slower than a proprietary protocol, but it is easy to read and learn, it is a standard available on almost every device on the planet and you do not even notice the difference in speed. Who cares about proprietary protocols? I certainly don't.

Sunday, June 04, 2006


This morning I checked on the ObjectWeb website the official number of downloads of the Funambol project and I saw we passed 500,000. Wow. Half a million downloads. Sometimes it is hard to believe even for me. In particular, when I see that more than half of them came in this year alone (and we have been alive and kicking since 2001...). The power of the community, a sign of good timing and a testament of mobile open source being the next big thing: I really believe we'll be the engine of the mobile Internet. Scary ;-)