My Foray into Forecasting

Google's recent OS announcement has got the 'net all tangled in excitement and debate.

This could be one of the most important days in computing history.

After reading this and other thoughts from Yan on the Chrome OS, I thought I'd throw my own into the growing mix. Though I'll refrain from making such grand prophetic remarks, I do support Google's innovation. In fact, I support any interface-building venture that deviates from the antiquated—dare I say obsolete—spatial-desktop metaphor.

Like many others, however, I have less confidence in the current plausibility of a pure thin-client approach. While the functional gap between web apps and desktop apps has shrunk over the past few years, it still remains a sizable leap. Without local storage, any application that deals with large files—image editors, as a common example—are bottlenecked by network transfer speeds. Web apps for further resource-hungry tasks like video editing, musical production, immersive gaming, etc. remain beyond the scope of current technology.

Worse, the thin-client approach today severely limits the functionality of the computer, even for a typical "light" user. Device usability becomes dependent on infrastructure. Even in the U.S. today, internet connections are not ubiquitous. The inability to, say, watch a video or listen to music on a train presents a real drawback. Of course, something like an iPod can fulfill this purpose—but why should consumers give up basic functionality that the cheapest of netbooks already have?

No, I don't believe the Chrome OS approach will soon replace traditional operating systems. That's not to say it won't be successful. Instead, it may become a popular parallel alternative to existing choices. It isn't unreasonable to assume that the aforementioned flaws just don't matter for most light personal computer usage. Google claims that the initial deployment will be in the netbook sphere—an ideal arena considering netbooks' typical usage.

In fact, if Google can make online storage transparent to the user, and maintain web apps' reliability and responsiveness, they may well take over the netbook OS space. The attractiveness of a few-second boot time is considerable; it's very well-suited to the netbook form factor and could become a defining feature. (Such a feature would allow netbooks to cut somewhat into the smartphone market, but I digress.)

But until we see significant progress in wireless infrastructure and network technology, I don't foresee Chrome OS spreading outside of netbooks, least of all to desktop machines. Capable desktop hardware is relatively cheap, magnifying the thin-client's artificial limitations in comparison to a traditional OS.

Whatever the case, fewer companies are as well-positioned for this venture as Google, with its considerable clout and sparkling brand. So here, I must disagree with Yan: open source or not, if Google can't pull this off, no one can.

Edit: I should note that I've used the term "thin client" improperly, here. A thin client relies, naturally, on a central server. Google's proposition does not exactly fit this description. Though Google does host popular apps on its servers, there's little reason to doubt that the Chrome OS will be able to run any web app. Thus the machines can be useful even in the absence of Google servers.