—”Zimbra has a testing harness thats in alpha which we will be making available to the public in the future, that measures performance on different actions within Zimbra. ”
Do you have any plan to open the automation way/tool for various browsers?

Another curious thing is: Chrome hasn’t open its’ interactional API, how did you achive to automate it for your benchmark?

Chrome’s main advantages have nothing to do with raw rendering speed.

You know how after a day of browsing with dozens and dozens of open tabs, your browser crawls to a near halt? I use Safari, and after a few hours of many-tab browsing I feel like I’m crawling the internet on a 14k modem. I have to restart Safari to get speed back. Essentially inefficiencies in memory management are rapidly compounding over time until the processor is doing 100x as much work as necessary to deal with user requests.

This is the biggest issue that Chrome was developed to address, and it’s where its biggest speed gains will be. Unless benchmarks are designed that simulate the memory issues of sustained many-tabbed browsing, they won’t come remotely near to showing the speed gains Chrome creates.

First off, Chrome is, as the article said, based off Webkit. Webkit is the open-source web rendering engine originally developed by Apple, on which Safari is based. Chrome is also open-sourced, and even now Chrome’s core technologies are making their way back into the Webkit repository, meaning that everything Google has done in Chrome is available in Webkit for Apple to use in their next version of Safari.

The second thing to understand is that Google simply has no interest in there being a ‘browser war,’ much less winning one. Their purpose is improve all web browsers so that they run web applications better, because web-based apps are what Google really wants to be about. Chrome’s main purpose was as a wake-up call to other browsers to take these new technologies seriously, or get replaced.

The thing I love about Google is that their definition of winning the browser war includes Safari and Firefox winning as well, and most importantly, web users everywhere winning.

@Peter Kasting
We used a customized OpenQA Selenium JavaScript test setup to calculate time taken rendering a page after clicking a particular button/link. This runs them all through the same set of ZWC tasks - such as logging in, composing and viewing messages, navigating around various folders, switching between our many apps, and even changing options. As soon as one task completes the next begins, thus total time for the entire rendering is a good measure.