Closed source is dead, open source is the way to innovation!

LXer Feature: 2-Jan-2008 Lately, some articles appeared which stated the open-source way of development didn't bring us any innovation. Jaron Lanier even goes further by saying closed source is the better approach to innovation. However, these people miss a lot of important points and facts about innovation, and therefore the conclusions they make are false. Having read a lot about innovation myself lately, I will try to show that the open-source way of doing things leads to more innovation, and more important, I will give some real life examples showing the closed-source inventions aren't that innovative at all, and pointing to some open-source inventions the other writers missed.

Most people (Mr. Lanier included) agree that openness has been the most important requirement for everything science reached the last few centuries. However, Lanier suggests not much innovation comes out of open-source because it lacks direction, though he calls it encapsulation. The problem is - innovation isn't defined; the word innovation itself lacks "encapsulation"; What is innovation and what not?

Nonetheless, most people will agree innovation finally brings something new to 'humanity'. There are several ways to reach this goal, and most known are Research & Development. Normally, those two are put together like they're one, but upon closer inspection they are more different than one might think. Basically, when trying to innovate, the 'cheapest' and fastest way is by means of development without research. Research is expensive and lasts long. However, if you take a plethora of existing ideas, choose some of them and combine them in new ways, you have a fast and cheap way to innovate. Note by the way these existing ideas have to be open-source because otherwise they can't be used to innovate. Also, if they are open source but patented, they cannot be used to innovate either. For example, while the patents on Watts' steam engine lasted, the basic efficiency and design remained the same. Only after the patent expired, third parties were able to combine the earlier patented idea of the steam engine and combine it with other ideas to make it more efficient, and only then the industrial revolution became possible. Speaking about combining existing ideas; the lion's share of innovation is done in this way.

Sometimes however, out of this plethora of existing ideas the right idea isn't there. Yet. That's when research is needed. A good example of this process is the CD-player. Back in the '50s of the previous century, almost every part needed to make a CD-player existed and worked yet. However, the fault ratio of the received bits was too high to make the CD-player work. That's when research was needed, and finally some mathematicians came up with error detection codes and even error correction codes, and the CD-player became possible. You could say, and this surely goes for the field of code-theory in maths, there was an 'outburst of radical mathematical creativity' which lead to the innovation of the CD-player.

So by now, I think one understands innovation doesn't have to be radical. What strikes me about Jaron Lanier's article is he blames the lack of direction in the open-source model for - what he assumes - the lack of innovation. I heard that before, but in a slightly different version. It was Theo de Raadt who said Linux is crap because a lack of focus. Also, there were persons who blamed the current mess at Gentoo on the lack of a social contract - , like Debian has - providing direction. Both the OpenBSD and Debian project on the other hand, do have focus. But are they innovative? While not radical innovations, there are still innovations in OpenBSD, but ask to the average geek and he (the average geek is a 'he') will answer there's more innovation in Linux than in OpenBSD. As a sidenote, the people at Debian and OpenBSD won't feel bad someone rates their projects as 'not innovative', because that isn't their goal in first place.

So, let's see how this fits in with Lanier's notion of innovation. First, there's the question: "Why did the Linux community didn't come up with the iPhone?" That one is simple to answer: The Linux-community doesn't make hardware. We also have to note, the iPhone isn't that innovative at all, only its marketing is. Apple is good at trendwatching and marketing, not at research or radical innovation. Rather, their innovation is conservative. They pick a product which sells to a small public only but isn't that popular yet, and than make it suitable for the masses. It's what they did with the flash-based iPod; I remember buying my first flash-based MP3-player in 2004 from the Chinese brand Apacer before Apple even shipped flash based MP3-players! Now, Apacer wanted to make a cheap MP3 player, but that was difficult because FLASH RAM was really expensive back then. That's when they had to be innovative, apart from the on/off button they designed the device with only one button for all functions , and they made the thing out of cheap materials. Sure, it was not fancy or robust, but it didn't let me down the first two years. FLASH-based MP3 players were scarce, and it was largely because I was too late to buy a "Discman" I found myself to be one of the 'trendsetters'. The same goes for the iPhone. The parts were available longer, for example the touchscreen comes from a company in Germany and most other parts come from factories in Japan. Moreover, LG blamed Apple from stealing the design and idea for the iPhone from its Prada phone it showed the public three months earlier.

So let's ask another question: Why doesn't closed software run on the OLPC? Apart from the designers deciding it would run Linux, the designers also couldn't have modified the closed software to run on the OLPC, but they would have had to rely on the companies providing this closed source software. In my opinion, the mesh-network of the OLPC's is a radical innovation. However, closed software doesn't support it. Microsoft is trying hard to alter XP in such a way it runs on the OLPC, but since Microsoft normally doesn't write Windows drivers they're not so good at it and it will take at least six months untill Windows runs on the OLPC. Also, the rather innovative Asus Eee ran Linux before Asus decided to also ship it with Windows.

Back to the iPhone once more. Though Apple may make some innovative hardware, I remember the touchwheel on the iPod - why does the iPhone lack GPS? Even Google and Microsoft know environment-aware computing is the future. For example, if my phone had GPS, my phone could know from looking in the address book where my friends are. And if they are near, you could design a game in which I have to press a button on my phone to virtually 'shoot' my friends if they enter the same room. Or you could devise a game, in which three persons using their phone 'trap' a fourth in a triangle, without even seeing each other. Those would be innovative games. But Apple doesn't have them.

Trolltech, on the other hand, does. They made a new open environment called QTopia". Therefore, people outside the "encapsulation" of the company can implement their innovative ideas. However, that very same people with their innovative ideas can't implement their ideas on the iPhone. That must be the reason Google came with Android; and Nokia - the biggest stakeholder in Symbian - is interested in a more open platform like QTopia; there's money to earn if you let third parties develop applications for your platform. An example of environment-aware computing would be your smartphone showing you the route to the nearest photo-printing shops after you made a picture with it. You don't have to know where you are at that moment, or where you are going too, environment-aware computing takes care of it. And if you're hungry, it will take you to the nearest pizza-restaurant with the same ease. If that shop was smart enough to buy localized Google-adds of course.

Then there's the argument Google uses closed source for its page-rank algoritm. There are two remarks to be made: Google wouldn't exist without Linux in first place; and maybe that page-rank would even be better if smart people could expand it if it was open. Just look at the number of Facebook-expansions to see how open development can benefit a website. OK, Flash may be the result of an closed development model, but it's meant to earn money for Adobe, not to be innovative. I mean, if Adobe wanted to make innovative software instead of earning as much money as possible, they might have chosen an open development model instead. On the other hand, flash _is_ innovative, but in such a way it tries to lock people in, because that's when money can be made. Probably Flash was one of the latest closed-source inventions anyway. Microsoft is far too busy keeping IE up with Firefox and Silverlight with Flash and telling people how good Vista is to do some real innovation*. I understood GIMP, Inkscape and Scribus are becoming that good in a way a lot of people won't need Photoshop, Illustrator and Framemaker anymore in the future.
*Update: Microsoft now even started remaking Emacs! QED, if you'd ask me.

As a sidenote, the LiveCD/USB-stick is a quite innovative open-source concept, and LiveCD's are used to repair Windows once it fails. Where Microsoft didn't innovate such a CD full of 'repair/recovertools' in more than 15 years, there are several out there based on free software now. Since open-source comes is projects with and without commercial interest, those without can offer innovative features big companies can't offer, like the automatic commercial skipping tool MythTV offers. Apart from the mentioned Microsoft and Adobe, and the peculiar closed-as-the-grave CAD software companies, almost all other software companies have embraced open source by now. Also, governments start asking for open source. The Dutch government has an "exchange"-website at which authorities can exchange both their open-source experiences, but also plugins they made. Rather open-source innovative projects can be found over there, like a program to migrate text-only municipality maps data to a database, or to couple GoogleMaps to Wiki's. That software might not have existed in a closed world, because there are many steps where it might have gone wrong.

First, it would be possible the company didn't know about the customers demands, then it could be the software wouldn't meet the customer demands, it could be too expensive, etc. With the open model however, if one municipality designs a piece of software, chances are far bigger it will meet the needs of other municipalities and the other municipality is aware of the demands of municipalities in general, than if the municipality went to a closed source software company. Moreover, the municipalities can support each other - which is much cheaper than buying support, and append to the sotfware if it doesn't meet their needs. At this point, it seems (some) authorities understand it may be cheaper to do software development yourself than outsource it, but only if you cooperate with other customers. For those authorities, that's a rather new innovative way of working.

I agree open source programs have catched up with closed source programs in the past, but right now they're not longer just mimicking but also carefully exploring the undiscovered road ahead, and if you're up to date you know open source programs have features closed software lacks.

OK, open source innovation might be chaotic because it lacks direction. There's no PR-department at 'the Linux community office' to ask what the current innovations are, what's going on and what's new. Instead of being lazy you have to find it out for yourself. Even someone who spends a part of the day reading news stories about open source software, finds at least ten innovative new programs he has never heard of in a list of 100 recommended open source programs (that's what happened to me today). No, I'm afraid people who think open source doesn't bring innovation and don't see how the closed source model hinders innovation just failed their trendwatcher exam. Wake up, it's 2008!