Last week I discussed Google’s Chrome Web browser and I got some great feedback and some less so. On the great side was an old colleague from Novell who said it was "by far the most interesting and informed … commentary I have read on Chrome." (I swear I didn’t pay for that.)

On the "less so" side was "Alex" who posted in the Gearhead forum "Your analysis is very basic, but fairly good." Thanks. I think.

Alex added: "I think that Chrome, at least for the first year of its life, is mainly intended for home users." Sorry Alex, but Google’s goal is much bigger than targeting the home consumer market. It is trying to build a platform for running applications delivered by the Web that relegates the operating system to a supporting role (Microsoft should be nervous).

Just think about it. When you can eventually run software as sophisticated as today's personal productivity applications by downloading them through your browser with your data stored in the "cloud", the prevailing user computing model that has been in place for 30 years is no more. Paradigm shift is such an over-used phrase, but that's what Google is trying to achieve.

The deep tech that underlies Chrome that I didn’t have space to go into in my "very basic" discussion is what will drive this new model. For example, Chrome uses multiple processes (one for each tab) instead of multiple threads, which provides better performance and memory management, and should prevent lockups and complete browser crashes due to the contents of a tab crashing. Chrome also has a JavaScript engine that is supposed to be significantly faster than the one in Firefox and much faster than Internet Explorer.

As to overall performance, there have been some interesting, albeit preliminary, tests. For example, a posting titled "Burning Chrome, Screaming Firefox, Lame IE" on the Open-Xchange blog concluded: "Google Chrome does not quite match the performance of Firefox 3, but in numerous tasks performed faster than Windows Explorer 7 … Google has delivered with Chrome a technically up-to-date Web browser, which performs nicely with demanding AJAX applications".

If you want more information about what's under the Chrome hood, check out Google's Chrome comic book (really!) and an interesting Wired article on the history of Chrome.

So, the proverbial bottom line: Don't dismiss Chrome (at least not yet) and don't classify it as another consumer-oriented browser. If Google can pull the browser rabbit out of the beta release hat, then much of what we do with PCs could change significantly. At the very least it could lead to new and improved architectures in other browsers, which would be a good thing in and of itself.

Speaking of browsers, a reader who had just switched to Firefox 3 wrote that when Firefox was running "every couple of minutes there was a flurry of hard drive activity, lasting maybe 10 or 15 seconds … I didn't see any Internet activity taking place, so I'm not sure what was happening, but I didn't like the idea that my drive might be scanned periodically without my knowledge or permission."

Interesting. I queried my friends on the-list-that-shall-be-nameless and a suggestion was made that as Firefox 3 switched from using .ini and .xml files for storing data to using SQLite, this might explain the disk access as housekeeping.

The absence of Internet activity may not be correct though. Both browsers and their myriad extensions are always calling back to their motherships for one reason or another, which will probably also trigger disk accesses. Mozilla recently published an article about Firefox making "unrequested" connections that is worth reading.

The reader has since gone back to "the 'old' 2.0 version of Firefox, which doesn't seem to need all that extra dusting and cleaning!" He added: "I also drive a 19-year-old Volvo station wagon that gets around 24MPG)."

Gibbs gets less than 24MPG in Ventura, Calif. Of course that is gallons of cabernet. Your mileage to gearhead@gibbs.com.