Oracle & Sun – The Open Source Lesson

As many of us know, last week Oracle filed suit against Google regarding what it believes to be a series of patent infringements by the search and mobile phone giant. While many are angry at Oracle, sensing an attack of the open source community, I’d like to take a different look at what has gone on. A bit of history and what the lesson could be for open source and our computing ecosystem at large.

A Quick Review …

For anyone that knows Oracle, it is no surprise that last week it filed a patent infringement law suit against Google. After unsuccessful licensing discussion with Sun, Google had decided to build Dalvik for its Android phone platform. Likely, Google did not believe it would run afoul of Sun’s Java patents, given Sun was not an ardent litigator of software patents (aside from its late nineties offensive with Microsoft). The curve ball: at the time Google made this decision, it had no idea that Oracle would one day own those patents.

So how did this happen? I believe the stage was set in the 1980’s with Sun’s nascent DNA and with free software evangelist Richard Stallman (and here).

Richard Stallman and Free Software …

In 1987 we cajoled Richard Stallman to come give a talk at our small computer science department at The College of William & Mary. Richard agreed on the basis that we would pay his transportation and feed him Chesapeake Bay blue crabs. With brown paper bags cut flat and spread open on a card table, the free software visionary proceeded to speak with us while breaking open crabs. Of course, Richard spoke of his beliefs on free software – how software should be treated from an intellectual property perspective. Yet what I want to share with you is a unique exchange during that meeting in 1987. In a small classroom with about forty of us, Richard sat cracking crabs with his long hair precariously close to crab claws.

A juxtaposed clean cut professor interrupted Richard’s commentary, “If it wasn’t for Lotus Corporation, we would not have products like Lotus 1-2-3” (Lotus 1-2-3 was a highly successful commercial spreadsheet program that boosted Windows PC sales. The professor was making the case that it was because of large companies like Lotus that a significant spreadsheet product like Lotus 1-2-3 could come into existence). Crab mallet in hand, Richard continued splitting crab legs and said, “I could write Lotus 1-2-3 in six months. The reason we can have 1-2-3 is not because of Lotus, but rather because hardware capability has gotten to the point that we can write software that performs such activities.” To my 21 year old ears, this was a profound comment – that software was the result of intelligent and motivated individuals, not the product of companies.

Sun – the early days …

Sun’s history is unique in the computing field. Bill Joy was a powerhouse programmer out of Berkeley, having authored significant parts of BSD Unix to include vi, csh, NFS. It was the first UNIX to have TCP/IP built in for networking. Bill Joy started Sun with a few other folks in the early eighties. With the combination of BSD Unix and powerful computing hardware, one could tackle heavy computing class problems with a Sun workstation or server at one’s desk. I began that journey programming Sun 2’s in the mid eighties. Through this whole debacle with Oracle, I’ve been periodically reading James Gosling’s blog. He has been with Sun since the mid eighties so many of his “wayback machine” references ring a bell. Sun was always good at hardware and operating systems. I spent most of my coding career building software on UNIX boxes – System V, NeXTOS, SunOS, Solaris, et cetera. Sun’s OS was exceptionally stable, but all kinds of service offerings overtop of the core OS were a land of near misses. The NeWS environment, printing infrastructure, Java IDE’s, Tooltalk in the mid nineties. So much technology above the OS never got traction. As a developer I learned early on not expect a Sun offering to be supported long after it was made available.

Deeply involved in programming an early analytics system built on the Sun platform, I remember a meeting of about a half dozen of us with Bill Joy, who had stopped in to visit our project. At the time Sun had adopted X Windows (post NeWS, et cetera) and we were having tenacious issues with colormaps. Sun had never been much of a graphics company (that honor was left for UNIX workstation competitor, Silicon Graphics). Bill went on a steaming rant about how MIT guys were hacking the X distribution to make some of the graphics work for games and how he no choice but to put up with the situation. When it got much beyond operating the hardware, Sun fell short time and again with its product offerings in spite of good ideas and the best of intentions. Creating sustainable business cases for technology ideas above the hardware seemed elusive to Sun.

With that said, I started programming Java systems in 1997. Java was a language then. It later became a platform which is nearly akin to an operating system, being a full up operational ecosystem. It was Sun’s third burgeoning success. Its first being the creation and sale of analytic workstations and servers in the early eighties and nineties, mostly to scientific and engineering communities. Second, was the onslaught of web servers and IT servers for the backroom, before Linux took a chunk out of that business. But Sun never really did software, I had lived it for over a decade when Java arrived. For years I owned Sun stock because I believed in the company, but I was flummoxed at how Sun could not monetize something as profound to the Internet and IT environments as Java became (and all its constituent pieces).

Which brings us to Open Source, Java and the Oracle purchase …

For the last decade I wondered who would buy Sun. In my opinion Sun needed to be bought, lest it go the way of Digital Equipment Corp (DEC) – a company of bright engineers with good product that ended up with no market and its constituent pieces sold off. My fantasy choice for a Sun suitor was Apple – they appreciate good hardware and operating systems. Apple’s market was front end computing, yet with difficulties in getting IT penetration. Apple knew how to product manage user facing computing products; I saw Sun and Apple as a dovetail marriage made in heaven. Obviously, I was the only one.

So instead it was Oracle. What did they actually buy for $5.6 billion? The engineering talent is fleeing at a rate that makes the Jews leaving Egypt in biblical times seem lackadaisical. Along with the hardware business, Oracle purchased a big chunk of “open source” – OpenSolaris, Java, OpenOffice and in the same traunch, MySQL from MySQL AB which was seeing heavy involvement from Sun. James Gosling likes to say that money is Oracle’s only metric. While Sun wasn’t necessarily focused on the “only metric” as money, they were struggling. Money is the primary metric of most large companies. Sun was different, possibly because of its inability over the years to figure out how to monetize its employees’ creations.

Ultimately, Oracle sees open source both as a competitor and a financial opportunity. A significant portion of Sun’s value was the open source it was shepherding. As much as Sun in its early days was not big on software patents, its Java patents were key to Oracle’s purchase decision. And Sun’s CEO and Board knew that. Sun was in precarious waters to survive as a company. The offering from Oracle was too good to pass up and the deal was done. As some predicted, OpenSolaris has been shelved. MySQL, an arguable competitor to Oracle, is now in its grasp – a coy move which slithered around Federal anti-trust review. Android is a big enough financial target to fire up litigation and go reap some rewards. What did Oracle ultimately want from Sun? It seems obvious. At the end of the day Sun sold it all to Oracle. To wit, even for Sun … and I hate to disagree with James Gosling in this regard … in the end money was the concern for Sun as it sold to LPOD (“Larry Prince of Darkness”).

And the lesson is … ?

Richard Stallman’s pronouncement over crabs in 1987 was correct – software is created because hardware exists which allows smart people to explore and solve problems in new ways. He saw much of the issues with software and intellectual property coming 30+ years before we did – it is the curse of intelligent minds to see the future and sometimes be right. For open source to survive, it needs to be in the hands of the community, held by non-profit organizations at best. Especially when open source products reach critical mass, becoming a core part of the computing fabric, they need to be shepherded by the independent entity of a non-profit.

Ultimately, companies are in the business of making money. Even as much as I applaud Sun for their varied illustrious contributions to a better world of computing, for the many nights I spent compiling on its hardware and living a life of bliss on its operating systems (instead of a life of hell on Windows), Sun was in the business of being a for-profit business and this ultimately collides with the community altruism of open source and free software. As its parting gift, Sun taught us a practical lesson that Richard Stallman warned us about more than thirty years ago.

Post to Twitter Post to Delicious Post to Digg Post to Facebook Post to LinkedIn Post to StumbleUpon Post to Technorati

You must be logged in to post a comment.