conormulligan.com / blog

The folks over at Chromium Blog have posted a lengthy write-up on multi-process architecture and how it applies to Google Chrome. It’s a must-read if you’re interested in this kind of thing.

Wired.com has an extensive write-up on the people and reasoning behind Google Chrome. It’s an interesting read, and helps put the whole venture in context.

Google’s answer to the question no-one asked landed today with the launch of their Chrome web-browser. And it’s actually quite good. It’s light, very fast, very easy to use, and does have some features to love.

Getting the most attention is V8, a new JavaScript interpreter Google engineered from the ground up for raw speed. It’s noticeable too, pages with a lot of client-side code are demonstrably faster than in Firefox and Internet Explorer.

Other distinctions include a very tidy interface that looks great on Windows Vista (Chrome is also available on XP; Mac and Linux users will have to wait a bit longer to get their hands on the beta) and a nice model wherein each tab operates within it’s own process; if the tab content is slow or crashes, the rest of the browser remains unaffected. It even includes a task manager to view the memory and cpu usage of all active tab and plugin processes (in fact, if you have other browsers open, it will include their resource information along side Chrome’s).

It’s not perfect, and is clearly still in beta. While stable, many killer features are still missing, like bookmark management, RSS feed integration and a proper plugin model. Also, the lack of a dedicated search box is really annoying for longtime Firefox users, although this concern may fade with extended exposure (Chrome’s navigation bar doubles as a search box, using keywords and shortcuts to choose a search provider). On the other hand, it does automatically add search engines to it’s collection when you use them (go to amazon.com, and moments later amazon search is listed in your search engine providers).

Overall, Chrome is a very promising entry into the competitive browser market, and while obviously an early product missing some important features, it is stable and usable for those who don’t require some of the complexities associated with Firefox.

As part of its dramatic User-Interface overhaul, Windows Vista added support for high-definition 256×256 pixel application icons. Unfortunately, Visual Studio 2008’s lackluster icon editor offers no support for the new standard.

Thankfully, the Visual C++ Team Blog brings word of a free ‘Lite’ edition of the impressive IconWorkshop from Axialis Software. IconWorkshop Lite even integrates directly into Visual Studio 2008.

The software is available here, and the Visual C++ Team Blog announcement is here.

By default, Firefox 3 now blocks Javascript code from hiding the browser navigation bar. The idea behind this is simply to stop malicious popups from emulating a native OS window.

But, for advances users well aware of what constitutes a malicious popup and what it looks like, the ever-present navigation bar can be distracting. Fortunately, you can disabled it by altering a setting in about:config.

1. Type about:config into the navigation bar and press Enter (ignore the warning).
2. Type dom.disable_window_open_feature.location into the filter textbox. A single entry should remain.
3. Change the value field to false.

You can now close the about:config window; popups can now hide the navigation bar like in previous versions of Firefox. To restore the default setting, simply change the value back to true.