Google’s “Open” Philosophy and the Conspicuous Lack of Open-Source Search

by on December 24, 2009 · 3 comments

Thanks to Jim for providing a great analysis of Jonathan Rosenberg’s “The Meaning of Open” from Google’s Policy Blog.  I wanted to throw in my two cents without derailing the comments on Jim’s post.  I hope you’ll this new thread of discussion interesting.

While I enjoyed reading Rosenberg’s post and found myself nodding in agreement with many if not most of his points, it would have been nice if Rosenberg were a little less cheeky about this close/open symbiosis that is the real defining quality of Google.  Rather than dismissing the closed nature of Google’s search/ad business with these lines:

The search and advertising markets are already highly competitive with very low switching costs, so users and advertisers already have plenty of choice and are not locked in. Not to mention the fact that opening up these systems would allow people to “game” our algorithms to manipulate search and ads quality rankings, reducing our quality for everyone.

Both of these arguments have some merit as explanations for why Google’s search/ad business isn’t open-source or an “open system,” but neither serve as a reason to grant Google an exemption from Rosenberg’s “open systems win” credo.

Instead of prescribing that the rest of the world adopt total openness, Rosenberg could have taken a more nuanced position, leaving room for the kind of proprietary money-makers Google relies on and that we’re not likely to see disappear from the software world anytime soon, if ever.  This sort of model, one which harnesses the profit-making potential of closed systems while funding satellite projects that take advantage of the iterative, peer-reviewed process of  open-source development is fascinating and makes for a much more interesting conversation than Rosenberg’s simplistic open-only philosophy.

Still, I think Google needs some defending and their business model/philosophy deserves to be looked at for what it really is, not what it is presented it to be.

Unlike Microsoft and other primarily closed-source companies, Google has been shockingly friendly to the open-source community and Rosenberg is right to brag about the millions of lines of code they have contributed to the community and the variety of tools they have made available to developers.  We should applaud Google for efforts like the Data Liberation Front (love the name) which is working to make easy exits from web services the norm, rather than the exception to the rule.  Despite its other flaws, this essay is brimming with examples of products and practices that should please open-source developers.  So let me be the first to say, “Goodonya Google!”

We also shouldn’t begrudge Google for keeping Search/AdSense/AdWords closed as that $20 billion (and growing) in annual revenue pays for a lot of free web services and funds many of the big projects that Rosenberg mentions.  That closed system has also kept a lot of engineers well-paid and happily productive, not to mention a nearly innumerable collection of data centers humming along with staggering efficiency.  Perhaps search innovation has suffered because Google has locked away it’s secrets, but then again perhaps nothing like the quality of Google’s searches would exist today without its closed-source cash machine.

But that really is the question here, isn’t it?  If Rosenberg is right and “open systems win,” why is there no (viable) open-source alternative to challenge Google’s dominance of search?  Switching your search engine is nearly cost-free, so why haven’t we seen Richard Stallman launch the GNUGLE project yet?

It could be that in some rare instances, the cathedral method of software development really is better than the bazaar (see Eric S. Raymond for clarification).  It makes sense that when confronting a problem like search—one that can only be solved through combining tremendous capital investment in physical equipment with software development—that a well-funded, centrally-organized firm like Google is better than a loosely-organized community of developers.

It may also be that Google’s congregation of developers is so large, that it has become a cathedral and a bazaar at the same time.  Meaning that Google may have a large enough number of fast and relatively independent developers to mimic how an open-source community operates while at the same time retaining the coordination abilities and financial resources of a large firm.

Then again, as the old adage goes, “Linux wasn’t coded in a day.”  Perhaps it’s only a matter of time before we see Google fall to devotees of the philosophy it can’t quite live up to.

This speculation brings me to a much larger point, a point that Rosenberg might have made in his essay had he been more frank about Google’s open/closed hybrid approach.  Rather than pushing for a one-size-fits-all openness policy for all developers, Google should be pushing for the meta-openness that is a free market chock-full of competition—something we’ve never had and likely never will, but it’s still an ideal worth pursuing.

Taking this position will always be wiser than adopting Rosenberg’s impossible credo.  Freedom to compete allows for the merits of open and closed—or Google’s unique brand of both—to be determined through consumer choice, rather than political wrangling and influence peddling, which rarely if ever produces ideal (or even remotely acceptable) outcomes.

This brings me back to the most powerful element of Jim’s post—that we needn’t fear folks like Rosenberg who occupy the Googleplex.  Instead we should fear the lobbyist and lawyers that occupy Google’s New York Avenue “Lobbyplex” in DC.  Despite the rhetoric about openness, it’s no doubt that Google’s policy pushers are working for one thing—and it’s not the not the progress of all mankind, or even the open-software/free-software movement—they’re working for the best interests of Google.

  • http://blaise.ca/ balleyne

    I think part of the problem with free software competitors to Google is that it's very hard to switch to a search engine that's under development. I can use an OpenMoko phone, which means it takes a few minutes to boot up, and I can't handle call switching to text multiple people at the same time, and make those tradeoffs. I can carry a second phone around (as I did at first), or just figure out how to make it work.

    But to switch to Wikia? Well, I tried using it when they were trying to build a search engine… and it was great to see it under development.. but if you're researching something, you're not going to say “oh, yes, I'll take sub-par results today because I want to support the development.” I mean, if you were looking up someone in the phone book, you wouldn't want to use a phone book that was incomplete.

    It's just tough to switch away from Google until someone else offers comparable results, but it's tough for someone else to offer comparable results if people aren't using/testing the search engine. The real switching costs are during development…

  • caroljns8

    It may also be that Google’s congregation of developers is so large, that it has become a cathedral and a bazaar at the same time. Meaning that Google may have a large enough number of fast and relatively independent developers to mimic how an open-source community.
    Mio Navman Spirit S300

  • caroljns8

    It may also be that Google’s congregation of developers is so large, that it has become a cathedral and a bazaar at the same time. Meaning that Google may have a large enough number of fast and relatively independent developers to mimic how an open-source community.
    Mio Navman Spirit S300

Previous post:

Next post: