Mobile

Critics Poke Holes in Android vs. iPhone Browser Test

The browser in Google’s Android mobile operating system is more than 50 percent faster than the browser found in Apple’s iOS, according to company Blaze.io.

Blaze tested the embedded browsers in Android 2.3 (aka “Gingerbread”) and iOS 4.3. These were WebView and UIWebView, respectively.

The tests were conducted against websites of Fortune 1,000 corporations.

Also among Blaze’s findings is that although both Google and Apple claim they’ve sped up JavaScript in Android 2.3 and the browser in iOS 4.3, respectively, it had no impact on browser speed.

However, shortly after the results were publicized, information surfaced suggesting the tests could be flawed.

Details of the Tests

Blaze used an iPhone 4 running iOS 4.3 and a Google Nexus S smartphone running Android 2.3 for its tests.

“We chose the Nexus S since it’s the poster boy for Android 2.3,” Guy Podjarny, Blaze’s chief technical officer, told LinuxInsider.

The methodology Blaze outlined appears impressive. It loaded Web pages from the sites of the Fortune 1,000 companies three times on each device, on different days, and used the median load times as a basis for comparison.

Results with speeds greater than 40 seconds or less than 400 milliseconds were weeded out because these indicate network and server errors, Blaze said.

The tests were conducted over WiFi in an area with good reception. They were conducted at night to eliminate noise and to maintain consistent results.

Measurements were conducted using a custom app developed by Blaze that uses the platform’s embedded browser. This loads a page on demand and measures how long that takes. Load times were calculated using the “Document Complete” callback from the browser, which Blaze contends is a standard way of measuring a Web page’s load time.

To distinguish mobile sites from non-mobile ones, Blaze loaded the same 1,000 pages through Internet Explorer 8 and compared the number of resources required to load the page on the iPhone versus the number required for IE 8.

If the desktop browser required 30 or more additional resources, the site was designated as mobile, meaning it had a customized version for mobile device access.

The Test Results

The tests showed that the Android browser was 52 percent faster on average than the iOS browser. Also, Android was faster on 84 percent of the sites tested.

Blaze also found that, although Apple and Google had optimized the JavaScript engines in iOS 4.3 and Android 2.3 and boasted of the resulting performance improvements, browsing speed wasn’t really different when it came to loading Web pages.

Blaze used real websites in its tests because custom-created benchmarks, such as the SunSpider JavaScript benchmark that’s normally used, don’t reflect the actual user experience.

That’s to be expected, because the SunSpider benchmark tests the core JavaScript language only, according to its Web page.

However, when tested against mobile-specific sites, Android 2.3 was only 3 percent faster than Safari on iOS, Blaze said. Mobile sites tend to be smaller and lighter than regular Web pages.

Putting Out the Fire

However, the tests are perhaps not quite what they’re made out to be.

“I’m skeptical because I think Blaze have an ulterior motive in conducting tests, seeing that they optimize Web performance,” Mark Beccue, a senior analyst at ABI Research, told LinuxInsider.

The results shouldn’t be taken as set in stone, because the speed of native browsers on mobile devices is affected by the processing chips used, Webkit, and the network, Beccue pointed out. “The Android browser’s speed is particularly hardware-dependent because the operating system is licensed to different hardware vendors,” he added.

Finally, the Nexus S is not the best test platform because “it’s not a widely used phone,” Beccue cautioned.

Meanwhile, The Loop points out that using an embedded browser is not the same as using the official browser.

UIWebView, which Blaze tested, is based on Safari but didn’t receive any of the updates Safari did in iOS 4.3, The Loop claimed.

Blaze did not respond to follow-up calls to discuss these issues.

But Really, Who Cares?

Perhaps the difference in speed may not be all that important, suggested Rob Enderle, principal analyst at the Enderle Group.

“On this last cycle, other things have started to take precedence over speed — safety and the ability to take advantage of the new multi-core processors,” Enderle told LinuxInsider. “Once you realize that speed problems generally come from the network or plug-ins you may decide to focus more on other aspects of the browser, such as safety, in order to differentiate your product.”

Further, the browser is not a focus of Apple’s, Enderle pointed out.

“Apple feel the browser is something they have to have, but not something that differentiates the product positively, nor is it something that helps lock the customer into the Apple ecosystem,” Enderle stated.

For Google, on the other hand, the browser is “critical to what they provide, and their future is Web-based, so they invest a great deal in making their browser competitive,” Enderle remarked.

Apple did not respond to request for comment by press time. A Google spokesperson declined to comment.

1 Comment

Leave a Comment

Please sign in to post or reply to a comment. New users create a free account.

LinuxInsider Channels