The Interesting and Not So Interesting of Google's Chrome

I'll readily admit that explaining the idea behind the Google Chrome project via a comic book created by the genius Scott McCloud (whose book "Understanding Comics: The Invisible Art" should be required reading for anyone who does any kind of visually creative work) is great. Licensing it via Creative Commons is even better.

But I'm not sure I get the need for another Web browser. If anything, I see this as merely a way for Google to continue to drive traffic to Google properties and bring a some "innovation" to the table in the process. If anything, it gives them an opportunity to start displaying "Works Best in Google Chrome" buttons on all their Web properties, just at the point that we've begun to collectively move away from Web applications that only work in IE or Firefox.

I'm pleased that Google has chosen to use WebKit for the rendering engine, so if a page looks fine in Safari, it'll look fine in Chrome. However, they're using their own JavaScript engine, so Chrome becomes yet another Web browser on which you must test your applications.

Looking at the features of Chrome that Google itself highlights in the overview comic book, some of their innovations just don't seem compelling enough to warrant another browser.

• Stability: Safari, for me, is a very stable browser. Firefox and even Internet Explorer aren't half bad either. Browser crashes in this day and age are usually the result of a) out of control memory consumption by bad JavaScript, b) plug-in interaction gone haywire, or c) an error thrown at the operating system level. Chrome will address a) by having each tab be individually threaded and sandboxed, which will prevent one bad application in one tab from bringing down the browser — but that means applications can and will still crash, such as writing an email in GMail (the example used in the comic book). The plug-in interaction really can't be handled unless Chrome bans plug-ins, which is unlikely. Chrome itself is built on top of other operating systems (Windows, Mac, Linux), so it's still going to crash when there's an OS issue. No big win here, other than the individually threaded tabs.
• A Faster JavaScript Engine: Um, SquirrelFish? Tamarin? SquirrelFish is going to be part of WebKit, and Safari 4.0, and it's mighty fast. Firefox 3.1 will contain a new, very fast JavaScript engine. Processors get faster and so JavaScript processing on those processors gets faster. Google's new JavaScript engine (V8) may be faster, and it brings to the table some interesting compile-time optimizations. It's going to generate machine code rather than run interpretatively, but I wonder if this will only work if you use Gears and the GWT. If you're not required to use Gears +/or the GWT, then Chrome still have to compile and translate the JavaScript in to machine code when the page is loaded, generating a slowdown on page load times. At the end of the day, to the end user, will it really be remarkably faster, especially as SquirrelFish and the other JavaScript engines are open-source projects with lots of smart people working on them from all over the world? If V8 turns out to be a truly excellent JS engine, then it's nice that it's going to be available for other browser makers to use. Time and performance studies will show if it's useful and adaptable for others to use.
• Security: Sure, IE 6 is very insecure. IE 7 is better, and there are exploits available for Firefox and Safari. But those browsers are pretty dang secure compared to where we were a year or two years ago. Apple may not be the best when it comes to openness and fixing exploits on the same day they're discovered, but they're not terrible, and serious exploits are fairly uncommon. Phishing detectors are built in to Firefox and IE, though sadly not Safari, so that's nothing new. The sandboxing Chrome brings to the table is interesting, especially as plug-ins (a common source of browser crashes and Windows-based exploits) are going to exist outside and "above" the sandbox for each individual tab. I wonder, though, if plug-in makers (ie; Adobe) are going to have to rewrite their plug-ins so they work via this new kind of sandbox. Are there improvements here? Yes, in the sense of the double-sandboxing of plug-ins, and the simple sandboxing of each page so that keyloggers or other malware can't see what you're doing in a given tab. That's a nice win for consumers.
• User Interface: Um, Safari's interface? Firefox 3's interface? How much simpler can you get with a default UI and still be able to click through basic tasks in a browser? The "nine most recent or related pages" view by default? No thanks, I'll go where I want to go.
• Open Source: Well, WebKit, the core of Safari, is open source. The Mozilla project is open source. The final result may not be entirely open source, but how is this a big advantage or difference for Chrome?

I'll readily admit that the "Each tab gets its own thread/processing space" is rather nice. It'll make things work more smoothly. I'd imagine, however, that it wouldn't be too hard for this to be added to Safari, Firefox, or even IE. I also like the idea of the Task Manager for the browser, so you can see what resources are being consumed by each tab and kill tabs that are problematic. (Though I do find it interesting in their illustration of this in the comic book that they point to plug-ins (read: Flash) as the real problem with memory bloat and crashing Web browsers.) And Google wouldn't be able to make the Task Manager work inside of someone else's browser, so they needed to write their own to accomplish that task.

I do appreciate the fact that Google is giving a wake-up call to browser makers. I do appreciate competition. I do appreciate the different approach they're bringing to the JavaScript engine and the double-sandboxing of tabs and plug-ins.

I'm still not convinced that this is a way of driving additional traffic to Google properties (well, of course it is) and that we won't start seeing "You must use Google Chrome to access GMail. Download now!" appear in the not-too-distant future. I also hope that this doesn't hurt the development of Gears for Safari, or Firefox, as I think that's an important project and the tools it brings to the table should be available across Web browsers.

I'm also not convinced that the average user will switch away from Internet Explorer. It's the default, it's what they're used to using, it acts in a way that they are used to, and users, quite simply, hate change. Google is a trusted brand, but that doesn't mean they're going to get my parents, or my co-workers, or the students at the University where I work to change from clicking on that "e" on their desktop to "launch the Internet." If they're going to take market share away from anyone, it's going to be Firefox and Safari, and, to a far lesser extent, Microsoft's dominant Web browser.