Apple

Ars Technica just did a review of one of the latest Android OS mobile phones, Motorola's Droid. However, in their haste to praise the phone as not just an improvement over previous Android phones, like the HTC G1, they highlighted a performance comparison to Apple's iPhone 3GS-- and declared the Droid the winner.

The problem?

The data says otherwise. They assigned the phones average times for loading a series of pages, on which the iPhone averaged 8.0 seconds and the Droid averaged 9.3. Then they drew a graph, noted "longer bars are better" and claimed the Droid's score of 9.3 was better than the iPhone's 8.0.

Even in the article it says that the Droid is only faster on some mobile-optimized pages, but you have to jump through some hoops to see that result, too:

The article states:

The results are pretty obvious, and match the results of the synthetic benchmarks. The Droid is slower than the 3GS in Javascript-heavy pages, but is as fast if not faster on the mobile-optimized pages.

The problem is that this is unsupported by the data unless you are only considering the m.digg.com page "mobile-optimized".

Here's a screenshot of my DragThing configuration. It's a collapsible drawer against the right hand side of the screen, with 11 clickable tabs that activate separate layers. The icon drawn in each layer's tab represents the category of apps I place there: Home, Palm, Internet, Productivity, Browsers, Communications, Utilities, Games, Moviemaking, QuckTime, and Graphics.

Clicking the active tab hides the dock, leaving only the tabs. When inactive it also becomes semitransparent.

Each individual item is clickable to launch an app, but can also be used to reveal the target item's location,