There are already a great many blog posts on Google’s announcement of its operating-system-in-progress; probably the most interesting one I’ve seen so far is from John Timmer at Ars Technica. Sample:
From a technological perspective, there appear to be some interesting aspects to rethinking the operating system. For one, by having an extremely narrow focus—bringing up a networking stack and browser as quickly as possible—Chrome OS has the ability to cut down on the hassles related to restarting and hibernating computers. And, aside from the browser, all of the key applications will reside online, security and other software updates won't happen on the computer itself, which should also improve the user experience. . . . More cryptically, Google also says that the users it views as its target market "don't want to spend hours configuring their computers to work with every new piece of hardware." That problem has plagued all OS makers, and none of them have solved it to the satisfaction of all users. It's possible that Google thinks it can do so, but given its general attitude (everyone should be happy with Web apps), it's equally possible that the company has decided that people simply don't need much in the way of peripherals.
And then near the end:
Will all of this work? Apple spent a couple of years trying to convince developers that they should be happy with Web apps, but it's clear that the arrival of native applications has been a significant driver of the iPhone's popularity. Palm appears to be trying something closer to Google's vision with the Pre, but Palm is also offering a native SDK, and it's too early to tell how well its reliance on online services will work out for users. At this stage, it's not even clear if the netbook market will have staying power once the economy picks back up.
We’ve seen already the convenience of web apps — access to the same data from anywhere you have am internet connection, and “pushed” upgrades that “just happen” — and we’ve seen some of the problems: catastrophic data loss (e.g. the ma.gnolia disaster), privacy concerns, lack of offline access, the limited feature sets of web apps in comparison to their desktop counterparts. Google’s approach to these problems seems to be to reassure us about the first, hope that we ignore the second, fix the third, and hope that convenience trumps the fourth. My guess is that ultimately they will succeed in all these endeavors, at least for a great many consumers.
Perhaps I'm a curmudgeon, but I cannot imagine why anyone would want to entrust all their private data to third party servers, let alone a company as powerful and ambitious as Google.
But then again, Google convinced some of the world's greatest libraries to hand over collections that cost untold billions of dollars and generations of painstaking reference work to accumulate.