Andrew Grossman, Heritage’s senior web editor, sent over the following on the just-released Google Desktop, and the looming battle between Google and MS. (To see more from Grossman, check out the Heritage policy weblog):
“Last week, Google, a company renowned for its search service, released the Google Desktop, a piece of software that lets users search through the materials stored on their own computers, from e-mail to Word files to Web pages that they have recently browsed. The Desktop is Google’s first major foray onto the desktop, and its release may mark the beginning of the end of Microsoft’s dominance of the desktop software market. Someone should tell the trustbusters in Washington and Brussels that their services are no longer needed, if they ever were.”
The Browser Wars, Reloaded
Back in the heady days of the dot-com boom, Netscape wunderkind Marc Andreesen declared his company’s browser the successor to Microsoft’s Windows platform. He said this, of course, before Netscape’s Navigator was crushed by Microsoft’s Internet Explorer. But Andreesen may yet be vindicated–though Netscape, which was swallowed by AOL years ago, won’t be the one to do it.
Until recently, little had changed in the world of Web browsing since Microsoft won the late-1990s Browser Wars. The Web itself was far from stagnant, but browser development slowed down considerably. This was understandable: Internet Explorer clobbered Navigator based largely on speed and stability. Beyond a certain point, neither can be improved much. And as Navigator, later Communicator, languished under AOL’s ownership, Microsoft had little reason to add capabilities to Internet Explorer.
But change was afoot at other levels. At first the concern of a fringe minority, Web standards soon came to occupy the interests of many on-line designers. Web standards are sets of rules that designers can apply to the code on their pages and “validate” their pages against. Standards replaced “tag soup” development that forced designers to create different versions of their pages for different browsers. Standard-compliant sites tend to be equally functional on all recent Web browsers and are often easier to develop and maintain than those using the “tag-soup” approach. With recent versions of XHTML and CSS–languages for defining the content on a Web page and its appearance, respectively–designers have been able to create sites with desktop-like functionality, but without the proprietary extensions, like ActiveX, or plug-ins that had been required in the past. As a result, the vast majority of Web sites are as usable in Mozilla, the open-source descendant of Netscape’s Navigator, as they are in Internet Explorer.
Mozilla’s slow development and recent popularity figures heavily today. AOL set the code for the Netscape browser free in 2000, but it wasn’t until around 2003 that non-programmers could comfortably use the resultant browser, Mozilla. What happened in the interim? Much of the original Netscape code was given the boot and replaced with cleaner-running, more portable versions. Web standards evangelists also became involved; in addition to supporting standards better than any other browser available, Mozilla created a few new ones itself. These newcomers, most notably one called XUL, veer into territory previously reserved for APIs, which are the special instructions that programmers use to interface with the operating system to, say, draw a window or italicize text. In short, Mozilla emerged from its cocoon no longer a mere browser for Web pages, but as a full-fledged platform on which developers could build applications. Still, Internet Explorer’s market share was such that few noticed.
But that’s changed. In recent months, a stripped-down Mozilla derivative called Firefox has begun to steal users from Internet Explorer, which had never lost share before. Mozilla has features such as pop-up blocking and tabbed browsing that were unavailable in Internet Explorer, but as a browser suite containing e-mail, newsgroups, and the kitchen sink, it was often poky. Firefox splits these other items into separate programs, making it a natural drop-in replacement for users dissatisfied with Internet Explorer. A rash of security breaches and the growing annoyance of ad-ware and spy-ware have convinced many users to make the switch. By the numbers, few have looked back.
Despite its recent success, Mozilla faces major challenges. Some time after AOL opened the Netscape browser code, it abandoned development of the browser completely to the nonprofit Mozilla Foundation. While the foundation is reasonably well-funded, core development of Mozilla has slowed. Mozilla needs a new major corporate benefactor if it is to continue its fast-paced evolution. If the rumors are to be believed, that may be Google.
Misplaced Files
Desktop search is one of those ideas that has been around forever that no one can get to work right. Ever used the file search function in Windows? It takes forever, just like searching for messages in Outlook. Finding things on the Internet is easy, but finding things on our own computers often seems impossible.
Microsoft has been talking about desktop search for some time and has promised to include a database-driven filesystem in Windows for over a decade. But that feature was recently bumped from Microsoft’s forthcoming Longhorn release of Windows. It probably will not be released before 2010, if ever.
How did Microsoft drop the ball? It banked on technology that only a geek could love: by replacing files and directories with a meta-data rich database, users could forget about where they were saving things and find files with simple and logical database queries. Simple, however, is relative; for this scheme to work, users would have to “tag” files with all sorts of identifying information, such as the file’s subject, its intended recipients, keywords, and so on. Anyone who has sat down at a friend’s computer and tried to find a specific document in a folder filled with files like “untitled 23” and “untiled 68” knows that the database filesystem would be dead on arrival.
The database filesystem is the opposite of the Google approach. The beauty of Google, both on and off the desktop, is its simplicity. There’s one blank. You type in a keyword or two when you search. It gives results that are generally sensible. Hiding a boggling complexity behind its clean interface, Google demands very little of its users.
So as Microsoft wrestled with its databases and pumped resources into the recent Windows XP update and Longhorn, Google leapfrogged the company with a desktop search tool based on the Google sensibility. The Google Desktop indexes more files and services than its competitors, doesn’t bog down the computer while it compiles its index, and ties its search results into the regular Google Web interface, which by now is familiar to most computer users. Having all of these implementation improvements at once is revolutionary in the same way that the original Macintosh was: It provides a long-awaited, useable implementation of ideas that have been floating about for years.
And it is also Google’s first major foray onto desktop computers.
Bundling It Together
Andreesen’s vision of the Web as a platform is slowly being vindicated. It is now routine to develop new applications for the Web, rather than for Windows, for a variety of reasons, including ease of deployment and cross-platform compatibility. Even complex services, such as customer relations and business operations, are now regularly implemented on the Web. And many companies are choosing to port their custom applications to the Web when they upgrade to ease future development. In short, the tasks that occupy computer users are increasingly being decoupled from Windows.
And at the same time that new Web standards, implemented in Mozilla, make it and the Web a better application platform than the old Netscape ever was, the Windows APIs are in flux with Microsoft’s plans to replace them completely with a more modern set. The result is that software makers are more eager than ever before to leverage the Web, as opposed to Windows, where they run the risk of making expensive investments to create software for a platform that few users may adopt.
Even now, for most users, only a few applications remain desktop-based. Word processing, spreadsheets, slide shows, and multimedia editing–that’s it. And even now, rudimentary implementations of all these tasks are available on the Web.
When users access only a small and shrinking number of applications outside of the browser, does it really pay off for them to upgrade to Microsoft’s latest version of the Windows operating system, or has the Web browser become their operating system for all practical purposes? This is no idle question; Microsoft launched Windows XP three years ago, and many users, especially corporations, are still reluctant to upgrade.
If it is as successful as the Google Search Bar for Internet Explorer, the Google desktop will be wildly popular. And given the current dearth of serious competitors and its attractive price (free), the Google Desktop may be even more popular than that.
Crucially, the Desktop software gives Google a toehold on users’ computers from which to introduce new features and capabilities. Google’s Search Bar updates itself without any user intervention. Will the Google Desktop software do the same, bringing users new functionality instantly, without the hassles of downloading and installation?
One current rumor has it that Google is developing a Web browser of its own, based on Mozilla. If Google chooses to deliver this through its Desktop software–perhaps as a way of integrating the Desktop with G-mail and other Web-based applications–it could steal a substantial chunk of the browser market overnight. Forget Internet Explorer; it is Microsoft’s Windows platform that is really at risk.
Many economists believe that something like this has been possible from the beginning. Microsoft, they say, has a contestable monopoly. It may be that the market for computer platforms is best served by a single dominant player, that the benefits of being able to use software and files on a variety of computers running the same system outweigh the costs of monopoly. But if the monopoly is contestable, which platform is dominant could change overnight. If Microsoft doesn’t take the threat of the Web-based applications seriously, it could quickly be yesterday’s market-leader.
Because of its enormous user-base, Google may be the biggest threat to Windows, but it is not the sole threat. Mozilla is open source, and any major player, from IBM to Sun, could take a shot at the Microsoft monopoly. In the end, no one company may be able to monetize Mozilla’s success, but many could profit from loosening Microsoft’s grip on desktop software.
After the tech bubble burst, Andreesen’s vision of the Web as a platform was widely ridiculed as naïve in the face of Microsoft’s dominance of desktop software. But technology is unpredictable, and Andreesen may have been only premature. This presents a fundamental difficulty for the antitrust authorities: The vagaries of technology and markets make even their most certain calculations–such as, for example, that Microsoft was an unstoppable monopoly that must be broken up or regulated–no more than guesses.
The lessons to take from the Google Desktop is that the markets have ways of dealing with monopoly and market dominance may be much weaker than it appears when markets are shifting and churning, as they always are. The need for antitrust authorities, then, is not apparent. If Google or some other company does succeed in winning the new platform wars, it would be no less of a mistake to go after them as monopolists than it was to go after Microsoft, which finds its position today increasingly fragile.
The Technology Liberation Front is the tech policy blog dedicated to keeping politicians' hands off the 'net and everything else related to technology.
Comments on this entry are closed.