Google Chrome vs. the World, Part 1: The Chrome Web Browser

A year ago, Google announced its Chrome OS project, ushering in a new front in its war on established OS vendors like Microsoft and Apple, one that is based on the web and web applications. (Read my Google Chrome OS Preview for the details.) This week, Google expanded on its plans, provided a new roadmap for its fledgling OS and padding out the supporting services that it feels will put it over the top.

I watched this week's Google event live on You Tube but tried to resist immediately writing about my reactions. As many of you know, I've been an ardent supporter of cloud computing since it became a term, and I moved my email, contacts, and calendar management out of the Outlook tar pits and up to Google services years ago and have never looked back. Too, I've championed the need for Microsoft--or, really, any platforms company--to simply turn its back on the past and start fresh with something new, built from the ground up for today's users' needs. (Historical footnote: When Microsoft began its NT project in 1989, that was the point. It's perhaps instructive to note that the company hasn't reset its platforms direction as completely since that time. That is, we're still running NT today.)

So I watch what Google is doing here, and a combined sense of dread and excitement sets in. On the one hand, I see Netscape reborn, promising to relegate Windows and its aging infrastructure to the dustbin of history. Only this time it's coming from a company, Google, that can actually make it happen, thanks to its ever-increasing cash hoard and a Terminator-like insistence on bringing its idea of the future to fruition, damn the costs. On the other hand, this is what I've always asked for. When Google says that no major new non-web applications have been created in, well, forever, they're right. When Google says that the future of computing is connected and mobile ... right again, and I've been saying that for years.

On the increasingly defensive Microsoft side of the fence, I see a disbelief that the ways of the past aren't working any more, when in fact they never really worked at all outside of a few core markets. I see a company that suddenly doesn't have the cheapest and most accessible solutions not know how to react to a competitor that is out-Microsofting it. And I see an unwillingness to change.

I'll get to some more direct Microsoft comparisons in a moment. But first, let's look at what Google announced this week, with an eye towards understanding how this impacts Microsoft, its products, and its customers. First up, the web browser.

Chrome web browser

At its event a year ago, Google said that Chrome had jumped to over 40 million active users, a figure I put in perspective by noting that Chrome accounted for barely 3 percent of all web browsers in use worldwide, behind even Safari. But those numbers have gone up dramatically since then. By mid-2010, Chrome had over 70 million active users, and as of now, the figure is a whopping 120 million users. That is simply incredible growth, about 300 percent since January of this year. Web browser market share--really, usage share--is hard to measure, but according to a median measurement of various market watchers, Chrome controls 11 percent of the browser market and is the number three browser overall, behind IE (52 percent) and Firefox (26 percent), but well ahead of Safari (5 percent) and Opera (2.5 percent).

Google's approach to Chrome is somewhat unique in that it developed the product as a "modern operating system for web applications" rather than as an application that runs under a traditional OS. This isn't just marketing speak: Chrome was actually designed, from the start, to incorporate OS features, features that will make it more compelling as the basis for an actual OS. Microsoft's approach, incidentally, is to make IE part of the OS and, in IE 9, to expose web applications to the user in ways that mimic traditional Windows applications. This makes sense--if you're the traditional OS vendor trying to maintain your market position--just as is Google's desire to wrest control of the browser--i.e. the delivery system for its own money-making services--from Microsoft (and Apple).

Google pushes three key aspect of Chrome: Speed, simplicity, and security. But let's be serious, Chrome is all about speed, and as any Chrome user will tell you, the primary reason they moved to Google's browser is that it just performs better. Give Microsoft a bit of credit for addressing this problem in IE 9, which indeed does feel about as fast as Chrome, on a new PC install. But the problem with IE 9 is that most users don't have clean PCs. And if you upgrade to IE 9 from your current IE version, it will bring along all the browser add-ons ("cruft") that were previously installed, slowing down the browser like before. This simply doesn't happen in Chrome: I've never seen this browser bog down because of add-ons, ever.

Go figure, but Google isn't sitting still in the speed department. At this week's event, the company showed off a number of performance-related tweaks that are coming to the browser. The Chrome Omnibox--its combined address bar and search box--is getting a new feature that auto-loads your most frequently-accessed web pages as you type. (What this means: If you access ESPN a lot, as in the demo, and simply type 'e' in the Omnibox, espn.com will load as you type.) Google also dramatically enhanced its PDF renderer, providing not just in-browser PDF display, but also near-instant rendering, regardless of document size.

In a bit of a jab at Microsoft and IE 9, Google is working on hardware acceleration for Chrome as well. Google is using WebGL to display web graphics using your PC's GPU. This is most certainly not as pervasive and powerful as Microsoft's all-in approach to hardware acceleration, but it wasn't by chance that Google chose an IE 9 Test Drive-like aquarium demo to make its point. (On the other hand, Google didn't bother to compare Chrome's performance here to that of other browsers, as Microsoft has done. For now, at least, Microsoft appears to have the edge in this area.)

Finally, Google is of course updating its JavaScript engine as well. Known as V8, this engine and Google's performance claims in the past, have triggered a weird browser sub-war, in which each company touts both the name and relative performance of their own JavaScript engines. Without getting into the pointless details--JavaScript performance is important, but only one aspect of overall browser performance--V8's performance has improved pretty linearly across new Chrome versions. And this week, Google announced an "enhancement" to V8 that comes with yet another name, Crankshaft. That Google chose to compare the speed of this technology to "IE from two years ago" (literally) is, perhaps telling. As with hardware acceleration, this is an area where Microsoft has improved things dramatically with IE 9.

Looking at the simplicity of Chrome, I've always appreciated this browser's low-key design. People talk about how various products "get out of the user's way" but there's a fine line between subtle and useless. Chrome somehow manages to be both subtle and useful, and really does get out of the way. Chrome's approach has been so successful, in fact, that Microsoft is aping it in IE 9, and to good effect. But let's give credit where it's due: Google was the first company to aspire to put the browser in the background, and make the web apps you use the focus. It's the right approach.

Google's simplicity--like that of IE 9's--is more than skin deep. The company highlighted some other areas in which the browser has been simplified. Software updates are seamless and automatic, something no other browser--and few applications--can claim. There are no modal dialogs, meaning small windows that pop-up and block your access to other browser windows. And this week, Google announced a new major area of simplicity: if you turn on Chrome Sync, you can sync every aspect of the browser--bookmarks, themes, extensions, and so on--to the cloud, so that your browser will always look and work the same way, no matter what PC or device you use. And this feature works: I'm on the Google Beta channel and have been using it for months. There's nothing like it on the IE side, certainly nothing as automatic, fast, and seamless.

As any Internet user will tell you, security is always a paramount concern. In Windows, Microsoft has worked for years to shore up the underlying platform and, in particular, its flagship browser. Those efforts have been largely successful, but there's still an underlying impression that Windows users somehow need to be extra careful online, and sales figures for unnecessary Internet security suites speak volumes to this perception. More important, perhaps, real threats do exist. And they continue to evolve.

Google highlighted three main areas of Chrome security this week: The seamless, automatic updates, its sandboxing of web applications/tabs, and a new feature called plug-in sandboxing. The automatic updates feature is nice, and cause Chrome to work like a web application: It's something that Google just takes care of for you, and one less thing that you have to worry about. Microsoft's approach here, of course, is to give the user choice. But the way it really works is that users have to think about security and wander over to Windows Update on their own occasionally to make sure everything is up to date.

Sandboxing is a traditional software development technique in which code is contained within a certain virtual space, or boundary, and is unable to affect--or infect--other similarly sandboxed code. Sandboxing occurs throughout Windows, both through innate functionality that has existed since the dawn of NT--the imperfect memory space sandboxing of applications, for example--and via functionality that's been tacked on over the years.

According to Google, only Chrome offers "true" sandboxing among browsers. It sandboxes web pages already, and via a new feature it will offer browser plug-in (what Microsoft calls add-on) sandboxing too. Making this work is difficult, since plug-in vendors have certain expectations around their ability to integrate with the browser. But Google says it is working with the major plug-in makers, including Adobe, to ensure that their plug-ins are sandboxed properly but still function. It's unclear if this feature is available in any form currently, but it appears to be more of a future direction than a currently shipping, widely-available feature. That said, plug-ins/add-ons is one area that Microsoft has ignored in IE 9, most likely because it simply has too much work to do elsewhere. But I think this is a mistake, and the IE add-on model needs an overhaul. (This isn't strictly related, but IE provides no way to remove add-ons directly, while add-ons can often magically just appear in the browser without the user's explicit consent.)

IE 9 vs. Chrome

When Microsoft revealed its plans for the IE 9 user experience back in September, there were some cries of Chrome copying--the clean new UI, and so on--but I think most understood why Microsoft is heading in this direction. To me, the biggest new features in IE 9 are those related to Windows 7 integration. Those users with both Windows 7 and IE 9 can mix and match local Windows applications with web apps on the taskbar, and switch between them as if they were the same thing because, logically, they really are the same thing. (Read more about this functionality in my review, Internet Explorer 9 Beta, Part 3: Windows 7 Integration.)

Of course, Chrome offers a similar ability to create application shortcuts for web sites and apps, though it's much more limited than the IE 9 offering. That is, yes, Chrome will let you create a shortcut to a web app--say, Gmail--on the Windows 7 taskbar. And yes, when you click that shortcut, Gmail launches as a pseudo-application in its own right, with no browser "chrome" and its own place in the ALT-TAB windows management hierarchy.

IE 9 goes quite a bit farther, adding support for Windows features like Jump Lists, Aero Snap, icon overlays, and thumbnail preview controls. These are all useful features, though web sites/apps will need to support them explicitly. But I'm not sure they make a difference. Remember IE 8 features like the Favorites Bar, Web Slices, and Accelerators? Probably not: Few people actually used these features, and it's possible that the IE 9 integration features--which, frankly, require a lot more work and foreknowledge on the user's part--will suffer from a similar fate.

And then there's the simple matter of supporting the stuff people really use. I happen to use Gmail, Google Calendar, and Picasa, so guess which browser application shortcuts work better for these particular web apps? It's not IE 9, and even if I chose to use IE 9 as my daily use browser, I'd still create application shortcuts for these services with Chrome. To a new generation of users, Google-friendly products are increasingly relevant. So it's neat that Jango has IE 9/Windows 7 integration. But who the heck uses Jango?

But there's more. While Microsoft has been glacially developing IE 9 over an incredibly, almost inexcusably long period, Google has delivered 8 Chrome versions in rapid succession, all released without fanfare. The situation with IE 9 now reminds me of that of Windows 2000, which went through a similar (if lengthier) development process, one that was so long that Microsoft just started adding features to the product over time, since the market was changing. So too, with IE 9: This week, in response to the FTC's call for Do Not Track functionality in browsers, Microsoft announced that, sure, they could add that to the ever-in-beta IE 9. What the heck. It's never coming out anyway.

Mark my words: If and when Google decides to add this kind of functionality to Chrome, it will just be a feature of the next version and it will ship in a few months, and then they'll move on to the next version and whatever new features that has. With Microsoft, it's like the white smoke going up before the introduction of a new pope. It's a major event for some reason, and it's a feature that wasn't even on the drawing boards two months ago.

But back to the theme of giving people what they want. In part two of this series, I'll look at Google's new Chrome Web Store, which provides a way for people to find web applications and "install" them in their browser. This is iTunes for web applications, people, and why Microsoft has nothing like this in IE (or in Windows for that matter) is unclear. But you can be sure that when they do announce something like this, it will be a big deal, it will come to many months--or years--too late, and it will be accompanied by another round of handwringing on the parts of people, like me, who don't understand why, how, or when Microsoft lost its mojo.

The Internet moves quickly. Google moves quickly with it. Microsoft, I think, wants to maintain a comfortable, almost stately, gait, one that made sense in, say, 1993. That's not going to cut it in this new world. I feel like I've been issuing this same warning for years, of course, and while I sometimes do see little signs of life--IE 9 really is high-quality, by the way, Windows Phone 7 is hugely innovative, and Windows 7 was competent to shut up the Mac fanatics for a while--I'm concerned about the big picture. I'm concerned about the perceptions. And I'm concerned about the execution.

Google quick, Microsoft slow. Google right, Microsoft wrong. Google win, Microsoft lose? We'll see.

More soon...