The Philippine Online Chronicles

This is the story of the Cathedral and the Bazaar.

It is always darkest before dawn.

The early 1990s saw the decline of the Mainframe business and the move to a PC-centric world.  It was a dangerous time for IBM, with 90 percent of its profits coming from the large servers and software business.[i] By 1993, IBM lost US$800 million dollars.[ii]

Apple was not in a better position.  The early 1990s saw the decline at Apple with the Apple II discontinued, with the Macintosh line aging and nothing to replace it with.  The company was taking a hit in market share and the world was fast moving towards the more affordable Windows centric world.  Apple’s fiscal year earnings by the end of 1995 dropped 47.6 percent.

As IBM fell and the subsequent defeat of Apple followed and with it a chapter had ended. Microsoft became the ultimate dominant force in the computing industry. Yet in the shadows was this phenomena called the Internet that was starting to be commercialized. It was a new chapter being written.

The nascent WorldWideWeb was set free in 1993 and there was a whole new frontier that was born.  Because of the Internet, there was this upstart operating system called Linux that only insiders and the experts knew about and boy, was it slowly changing the world.

In that brave new world, a company called Netscape built and sold Web Browsers and a suite of Internet tools.  It was born in 1993 and by August of 1995 had an initial public offering at USD7 per share and it would reach its all time high at USD75 per share in December of that year.^[iii] Microsoft announced that it would focus its attention on the Internet.^[iv] Internet Explorer 2 became a free download thus, beginning the first browser war.

The mid-1990s saw a massive push on the Internet and in technology in general. Aside from Netscape, Red Hat Software was born to take advantage of the Linux phenomena.  Yahoo was one of the first companies to take advantage of the Internet and its growing popularity.

Everything, it seemed, would return, but in changed form.

By the fall of 1997, Netscape would report a USD88 million quarterly loss.^[v]

The Cathedral and the Bazaar

Eric S. Raymond delivered a paper before Linux Kongress on the Cathedral and the Bazaar. It talked about a little software project he wrote and the two fundamental differences in software development. Raymond’s paper explained how development was done the "open source" way.

Raymond called the closed-source and propriety model of software development as the Cathedral.  The masters of the code would do their voodoo behind closed doors, away from preying eyes. It was (and still is) the way Microsoft wrote software. In many ways the Microsoft way was the Bill Gates way and his belief in how software should be written traces back to the days when Microsoft wrote and sold Basic (see part 1).

Microsoft for all the “evils” the world seems to attribute to it, was truly the first Software Company, as the world knows it. At its core, Microsoft is not really an Operating System Company, nor an office productivity company; it really is about writing development tools, writing software.

The way Microsoft does it is to hold on to the source. To develop everything in-house or to acquire it and this model was the antithesis of the hacker culture of that sprawled in the 1970s, that gave birth to the Internet and the Web and to things like Unix, and C and eventually Linux.

The Microsoft belief is that people who make software can write software and ought to profit from that hard work.  It was capitalism.

What of the Bazaar model?  It was all about Open Source.  It was a software development methodology and philosophy that was open to peer review. Meaning anyone had access to the source code, it was available online.  Anyone could download it.  Study it.  Anyone could make changes to the code, so long as they make they make the changes available under the same conditions as the original.  Anyone could turn that source material into a running piece of program.

The bazaar way traces its fundamental philosophy to Hacker culture of the 1970s.  It traces its linage to the Unix philosophy and it was influenced quite obviously by Stallman’s Free Software Movement (see part 2).

The bazaar way was community centric.  People itching to solve a particular problem, could gather together and work on it together and this sense of community of working towards a goal that powers Open Source.

Successful Open Source

In the last 20 years some of the most successful open source projects have been, Apache webserver, BSDs, Linux, GCC, Mozilla Firefox, MySQL, OpenSSH, OpenSSL, PHP, PostgreSQL, Python, Ruby, The Web, and Webkit browser.

These projects have spawned success in fields like High Performance Computing.  The commoditation of the PC market has enabled a drop in hardware prices to allow the cheap construction and deployment of clusters.  Driven by Linux, these machines are now focused to crunch more numbers and so affordable that you could make your own cluster for under USD2,500 dollars.

The success of Open Source has inspired others to share.  It gave rise to Wikipedia, and to Creative Commons licenses.

Linux and open source tools serve as building blocks for huge technology companies like Google that customize their software.  The Internet is driven by these open source tools, some have built and some purchased from a supplier.

The core of Mac OS X is similarly Open Source.

The success of Open Source has been to keep the Internet going.  Open Source make up keep components of infrastructure around the world.  It is to Linux that mobile phones like Android trace their linage. It is to WebKit that mobile web browsers on BlackBerry and iPhone owe their pedigree.

The failure of Linux

Hacker culture is all about scratching an itch.  They want to solve a particular problem and they do it. Is scratching an itch enough?

Twenty years later, this is the limit of Open Source.  This is the sad state of affairs.

In the Webserver business, 60 percent are Linux servers. However, it accounts for less than 2 percent of desktop computers in the world.

Linux accounts for 90 percent of the high performance computing business.

According to IDC this is the server market at the beginning of 2010. Windows Server still owns more than 41.6 percent of the market, with over 4 billion in revenue. Linux represents 14.7 percent of servers at a 1.9 billion business.

The argument that companies don’t trust Linux is a misnomer. If high performance computing with its demanding requirements rely on Linux, why shouldn’t ordinary companies do so?  Companies -- particularly the geniuses running the backends -- simply prefer running things they know: Windows.

Education about the stability and training on the use of Linux remains the biggest stumbling block to its complete and utter domination of the world.  How else to slowly migrate systems off Windows and into Linux servers?

In February 2010, Red Hat ($RHT) reported operating margin at 14.27 percent, with net profit margin at 11.94 percent with a market capitalization of 5 billion dollars (at the time of this writing).

Red Hat as a company is doing all right.  One would think that such growth would be phenomenal. The adoption of Linux in the enterprise while increasing is still slowly getting there and for a company that provides for middleware and support for a server operating system that’s not bad business at all.

On the other hand, Desktop Linux is almost non-existent.  While there are great strides made in making Linux desktops easy to use, easy to install as well, there are too few people who have taken advantage of it.  There has been little significant push to create a single desktop operating system to the PC in the same vein as Mac OS X is to the Mac. An argument could be made that Mac OS X is the Unix-like desktop operating system that Linux ought to have been.

On mobile space, this is where the battle is happening.  Great strides are happening with Android.  Where Linux failed in the desktop, would it then succeed in the next generation of devices?  Android driven tablet for example could be what Linux desktop failed to be: a computing abstraction for a whole new generation.

Did open source fail?

Open Source succeeded where it was strong.  It dominated the space where infrastructure is most needed.  It makes up much of what keeps the Internet going.  It has taken the high-end computing genre to a whole new level.  It is moving quickly into the mobile space.

There are undoubtedly limits to what open source is.  It is great for the small tools and specific tasks.  It is phenomenally; superior software and the community that exist around it comprise some of the smartest people on Earth.

Open Source failed where it needed to be mainstream.  There and continues to remain an intellectual gap between the geniuses who without doubt keep open source going.  The translation to mainstream hasn’t been that successful.

The Marriage of the Cathedral and the Bazaar

It goes back to Woz and the Homebrew computer club.  Just because Woz would share his schematics and how he built Apple I, doesn’t mean everyone had the technological skills, the time or the inclination to build their own version.   The same has happened with open source.  The data is out there to be mined, but that doesn’t mean that the people in the fringes could understand it.

Remarkably, the biggest answer as to how to translate this technology for the masses is through the marriage of the Cathedral and the Bazaar.

Open Source software powers the majority of the web hosting business and whole industry grew out of it.  Some of these businesses wrote their own code to interface with open source technology.  How many people would really bother to know that a hosting provider ran their hardware off Linux?

Mobile devices have married the open source Webkit technology to their proprietary mobile devices.  Blackberry is one. Apple built their software on top of Open Source, paired it with their own proprietary code and polished it.

Google has married their secret sauce search engine with ubiquitous commodity hardware and Linux operating system.

The success and failure of open source and how the world turned out to be, which is a marriage between the Cathedral and the Bazaar is deviant to what Richard Stallman thinks of what Free Software is and ought to be.

The success of open source has been when one company takes that piece of software and translates it for the world.  Remarkably, the middle ground is what makes the world go around.  This is why Openness isn't what it used to be.

Previously on Lintech’s “Reflections on Openness” series, Part 1: The mother of all demos and Part 2: In Anarchy We Trust.  The Success and Failure of Linux is part 3 of this series.

____

The Story of Tux, some rights reserved.

Top 500 Operating System family share, some rights reserved.

[i] L.V. Gerstner, Jr., Who Says Elephants Can’t Dance? (Harper) 42

[ii]Ibid.

[iii]W.G. Rohm, and R. Young, Under the Radar (coriolis) 67.

[iv]Ibid, 71.

[v]Ibid, 67