I like the idea of Google Chrome, a lot. But, then I like a lot of ideas, and then the reality turns out to be another matter entirely. Once in a great while, though, something comes along that lives up to its promise. Google Chrome lives up to its promise.
That promise, as I describe in The real reason Google is making Chrome, is to take out Microsoft Office and, in larger terms, replace the desktop application metaphor with a Web application one. I don’t think, however, that Google wants to get into the operating system business or replace Windows.
I do think that Chrome, once it moves to Linux, has the potential to be a big help for desktop Linux. If Google is successful with its Chrome scheme, then a side-effect will be to dwindle Windows’ market share.
For some basics on what’s what with Chrome, Barbara Krasnoff does a fine job of reviewing Chrome. What I did was to see if Chrome lives up, in practice, to its promises of faster, much faster, Web application, specifically JavaScript performance.
To do this, I installed the Chrome beta on a Gateway 503GR. This system uses a 3GHz Pentium IV CPU, 2GB of RAM, an ATI Radeon 250 graphics card, and a 300GB SATA (Serial Advanced Technology Attachment) hard drive. On this older PC, I was running XP SP3.
After I installed it, and imported my bookmarks from Firefox 3, I just roamed around the sites I visit multiple times a day such as ComputerWorld, NewsForge, and NetworkWorld. One of the things these sites have in common is JavaScript-based ads. With Firefox 3, these sites are often slow to initially load; with IE (Internet Explorer) 7, these sites are painfully slow to load; with Chrome, "Wow!"
"Wow!" is these sites exploding, instead of crawling, on my display. A quick run of the SunSpider JavaScript benchmark revealed that it wasn’t just my perception. The test showed that IE 7 had a result of 59,682.1 milliseconds; Firefox 3.01 came in with 11,267.1 milliseconds; and Chrome ripped off a remarkable 3,617.8 milliseconds.
This, I might add was on a PC that was state of the art for 2005. On a newer system, Chrome’s results would be even more impressive.
OK, that’s great, but what about JavaScript-based applications like Google’s own applications. To find out, I ran Gmail, Google Docs, and Google Calendar.
Guess what, these applications on Chrome blasted by the same applications working with the same data on Firefox and IE. It seemed to me that Chrome was running at least twice as fast as did on Firefox. I won’t even mention IE.
How fast is that really? Fast enough that, for the first time, I can see ordinary users using Web-based applications instead of desktop-based applications for their every day work. Microsoft Office look out.