> Try to add a powerful new feature to Windows, and casual users balk. One example Miller cited is multiple desktops, a feature in OS X and Linux, but still not built into Windows. Each time Microsoft conducted user tests on multiple desktops, casual users got confused, prompting the company to cut it.<p>The same feature's confusing to causal OS X and Linux users, too. Which is most definitively <i>not a problem</i>, because casual users won't be coming into contact with that feature, anyway. OS X is a single-desktop interface until you actively choose to create a second desktop. Most Linux setups behave roughly the same. On territory this well-trodden, it should have required precisely zero lateral thinking for Microsoft to figure out how to respond to that focus group result.<p>Similarly, when Apple added a tablet-inspired application launcher to OS X, they figured out a great way to make it both easy-to-access and non-disruptive, without displacing any existing workflows. The launcher's there and works well for users who want it, and is effectively non-existent for users who don't.<p>That anecdote from the article, if true, is yet another delightfully pithy illustration of Microsoft's fundamental difficulty with UX: They don't grasp the difference between "one size fits all" and "all fit one size".