The Register®

Original URL: http://www.theregister.co.uk/2010/07/07/javascript_and_browser_tests/

Android spanks Apple’s iOS 4 in JavaScript race

Firefox 4 sticks to slow lane

By Rik Myslewski in San Francisco

Posted in Applications, 7th July 2010 18:30 GMT

Watch Now : Virtual Machine Movement with Hyper-V

The JavaScript engine in Google's Android 2.2 running on a Nexus One phone soundly spanks Apple's iOS 4 incarnation running on an iPhone 4. Also, This Just In: Firefox is Still Slow.

The Android versus iOS test results were announced [1] by Ars Technica, while the snarky-but-credible verdict on the just released [2] beta of Firefox 4 is the result of tests conducted [3] by The GigaOM Network's WebWorkerDaily.

The Ars testing discovered that in pure JavaScript performance, there's really no contest between Android 2.2 and iOS 4. When running the industry-standard SunSpider [4] benchmark (which you can run here [5]), Android 2.2 was nearly twice as fast as Apple's offering.

iOS 4's comparative performance was even worse on Google's own V8 benchmark [6], which you can run here [7]. Ars found Android 2.2 was well over four times as fast. Of course, V8 is a Googly benchmark, but 4X is 4X.

As The Reg reported [8] last week, development tools firm Appcelerator recently surveyed US developers and learned that a higher percentage of them think that Android has a better long-term outlook than does iOS. The Ars tests, it may be argued, add weight to that viewpoint.

Mozilla's Firefox 4 beta, which you can download [9] now after last week's false alarm [10], was less embarassed by its competitors than was iOS 4 — but if Mozilla wants to catch up before Firefox 4 comes out of beta, it still has some work to do.

WebWorkerDaily ran both Mozilla's own Dromaeo [11] test suite - the "recommended" version of which you can run here [12] - and Futuremark's Peacekeeper [13] (run it here [14]). WWD chose these tests, they said, because both test suites are more comprehensive than SunSpider and the V8 benchmark.

When running the Dromaeo test suite, the first Firefox 4 beta was solidly twice as fast as Firefox 3.6. That's the good news for Mozilla fans. The bad news is that it's still only abut two-thirds as fast as Safari 5 or Chrome 5.

In the Peacekeeper benchmark, the Firefox 4 beta showed less improvement over Firefox 3.6, was a bit more than half as fast as Opera 10.6 and Chrome 5, and two-thirds as fast as Safari 5.

JavaScript-engine efficiency and test-suite performance speed are of course not the end-all and be-all of browser evaluation. Many a user will put up with a wee drop in performance if, as in the case of Firefox, there's a healthy add-on culture plus nifty UI and productivity features.

We're all easily seduced by speeds and feeds, and sometime forget to evaluate the entire software experience. That said, as we mentioned above: "4X is 4X". ®