oreilly.comSafari Books Online.Conferences.


Whence the Source: Untangling the Open Source/Free Software Debate

by Thomas Scoville

The battle for ideological control of computing's next wave is being waged between two factions. One seeks to increase the collective IQ of the software development community by loosening industry's grip on intellectual property. The other wants to do away with intellectual property altogether.

In one corner stands the Free Software Foundation, conceived, championed, and largely engineered by the impracticably messianic Richard Stallman. In the other, the Open Source Initiative, advanced most prominently by the cantankerous, opinionated, and confrontational Eric Raymond. Both leaders are longtime programmers. Both are significant contributors to Internet culture, each with his own high place on the heap of information-age theoreticians.

Both sides agree that distributing a computer program's source code -- the essential blueprints required to modify the program -- is a good thing. Both believe that empowering users with the opportunity to change and extend their software will have a positive, reinvigorating effect on software development at large.

That is where agreement ends. And considering that both sides have completely different ends in mind, that should come as no surprise.

The Free Software Foundation was founded on the appealingly utopian -- and perhaps quixotic -- notion that all information should be shared. The Free Software ideology focuses largely on traditional copyright, and what its adherents perceive as fundamental injustices contained therein. In this idealistic view, any contract that prohibits the free sharing of information forces the individual into an adversarial relationship with others, and thus diminishes us all as people. More plainly, we can no longer be friends when the law requires us to keep secrets. Good, useful thoughts -- with computer programs considered particularly functional concentrations of such -- are imprisoned by copyright; the resulting unequal access, primarily by those with money over those without, is fundamentally unjust.

Richard Stallman's Free Software Foundation seeks to redress this injustice -- within the scope of the computing world, anyway -- by replacement of the standard copyright agreement with the "copyleft", a somewhat romantic licensing scheme which, though it does not actually prohibit sale of FSF software or its products, capably skewers the monopolist's ambitions. It effectively guarantees that any programming project having benefited from FSF technical leverage can never be considered intellectual property.

Eric Raymond's Open Source Initiative (OSI), on the other hand, is galvanized by the rather more utilitarian revelation that users are much less clueless than previously imagined, and that enlistment of their cooperation -- by allowing them to contribute to the ongoing process of improving the software they use --results in much more useful, well-designed, robust creations.

Though perhaps a less epic position than Stallman's transcendentalism, it's nevertheless revolutionary in the software business. Historically, source code is the jewel in the crown of intellectual property, guarded at any cost. Source code theft has been at the center of any number of high-profile Silicon Valley litigations. The newfangled notion of allowing customers -- and competition -- free access to source is, to many software industrialists, more than a little threatening.

But the Open Source Initiative has staked its reputation on this proposition. In his extended essay The Cathedral and The Bazaar, Eric Raymond makes a compelling case that this continued liberation of users will precipitate a healthy decentralization of the software business. He asserts that giving users -- whether they're customers or otherwise interested parties -- access to source code will create a new world of distributed productivity and greatly expanded opportunity, especially for smaller software developers.

The OSI camp, obviously, does not have the same philosophical problems with unequal access to the superior systems that may result from this new, open development framework. They assert that sharing source is a good strategy for high-minded utopians and cutthroat capitalists alike.

In any other field of endeavor, there would be no contest; the fluffy bunny of utopian daydreams would meet the oncoming diesel locomotive of commerce, and it'd be all over but for the rabbit stew.

The problem is, you can't ignore Stallman.

Aside from being a fanatic with an unrealistic, uncompromising vision, Richard Stallman is also an uberhacker, having began his career long ago in one of the most formidable pockets of innovation in computerdom -- the MIT Artificial Intelligence Lab. Since the breakup of this digital Valhalla -- its best minds having been hired away by industry in the 80s -- he's put an incredible amount of his legendary life and programming talent into the Free Software Foundation's GNU project. GNU is a collection of programming tools and utilities, most conspicuously including a compiler, the essential tool for turning source code into a running program.

One of the programs largely made possible by GNU is Linux, the much-vaunted free operating system that looks these days like a good bet to bury Microsoft NT. Of all the Open Source success stories -- Netscape, Perl, Apache, sendmail, bind -- Linux is the one most likely to keep Bill Gates awake at night. One could easily make the case that Linux would not otherwise have been realized without GNU and Richard Stallman.

This circumstantial evidence makes it pretty easy to perceive Stallman's generous, virtuoso effort as the technical foundation of the movement. Throngs of Free Software Foundation enthusiasts do, and thus seem to implicitly accept his radically socialist ideology as the One True Philosophy of source code liberation.

But there's another problem: Stallman wasn't the first.

Years before he or Eric Raymond ever hit on the idea of liberating source code, the UNIX operating system was being developed at AT&T Bell Laboratories. As a government-regulated monopoly, AT&T was barred from competition in the computer industry. Though UNIX source code was not then "free" in either the FSF or OSI sense, it could be licensed at nominal cost.

Universities were among the first to take advantage. As a result, UNIX ended up in the hands of hundreds of collaborating academic programmers. In particular, the UNIX effort at the University of California at Berkeley spawned a West Coast hacker culture to rival Stallman's MIT cohort. Ultimately, the student programmers at Berkeley created their own variation of the operating system so potent that it became a major fork in the UNIX lineage -- the Berkeley Software Distribution, or BSD.

It is difficult to overestimate the role of BSD UNIX in modern computing. Not only did it beget many key features of all future versions of UNIX, but it was also under the BSD flag that UNIX met the Internet (though at the time it went by its more ancient name, Arpanet). Much of the most common system software surrounding the TCP/IP protocol was developed at UC Berkeley, and was introduced to the world as part of BSD.

In the years since, BSD has enjoyed not only a substantial commercial run, but has also found its way into a commerce-free distribution of its own, one to rival Linux. Though not as popular or mediagenic as Linux, FreeBSD can nevertheless be widely found on the machines of hobbyists, ISPs, and major corporations alike.

So the shared source collaboration concept had received significant validation long before Raymond or Stallman showed up on the field. That would make AT&T the unlikely mother of the movement, having quietly accomplished the feat with neither Stallman's righteous rhetoric nor Raymond's theoretical grandstanding.

So where does that leave the current debate? The BSD story might serve to place the Free Software/Open Source tension within a wider historical context. In that broader view, one might discern a pendulum effect swinging through the software industry, between open and closed, cooperation and competition. Perhaps the interval of rampant Gatesian monopolism between the blossom of BSD and the advent of Linux is just another periodic lilt of the paradigm dance. Now that its momentum has run its course, things are once again moving the other way, as evidenced by the surge of support for Open Source.

This interpretation certainly leaves open the possibility for both Stallman and Raymond to collect a measure of glory, both having played key roles at different points along the pendulum's path. Stallman's utopian reveries -- untenable as they may be -- may have been the only ideological kindling dry enough to maintain the flame of open collaboration during a time in which the software industry was water-logged in a proprietary bog. Put another way, if the Irish saved civilization from the Dark Ages, perhaps Stallman's Free Software Foundation saved Hacking from the Age of Microsoft. But now that the idea of Open Source no longer seems completely psychotic to the captains of the computing industry, perhaps Eric Raymond's Open Source Initiative will orchestrate the integration of the shared source collaborative framework into the industry's mainstream.

Much as the playing field seems tilted towards Raymond, however, there is still cause for caution. Stallman, after all, still has a point: greed, competition, and secrets really are, on some level, bad for us; they really do rob the industry of a certain amount of genius. The time and energy that is poured into the maintenance of proprietary frameworks could be much better spent elsewhere, like on real innovation. Raymond's version of the Open Source story proposes no remedy, and does not even speak to this danger in a compelling way.

And Raymond -- strong as his position may otherwise be -- still has something to prove. Never has the Open Source concept really been tested in a radically competitive arena like the computer industry has lately become. It's not clear that corporate hard-chargers will be able defer their attention from IPOs, quarterly reporting, and intercorporate knife-fighting long enough to recognize how the Open Source proposition may ultimately appeal to their self interests. After all, these aren't the Regents of the University of California we're talking about; these are hip-gunning cowboy capitalists. Ideological high-ground aside, it will be interesting to see if the Open Source concept holds up in the cross-fire. Whether it's ultimately about the beginning of a new cooperation or the end of intellectual capital, the Open Source alternative is going to have to withstand a lot of manhandling in the marketplace. It wouldn't be the first golden opportunity to be lost in the rough-and-tumble.

Sponsored by: