In the old days of Safari 1.x, Apple’s web browser for Mac OS X, there was a hidden WebKit preference value named WebKitInitialTimedLayoutDelay, which controlled how much time the browser would wait before it started to render the page. This was so that some content would have come in for the browser to render, so that you don’t have weird graphics glitches as the page and CSS loads. If you tuned this value lower, Safari would render the page as soon as it gets pieces of the page, leading to a perception of “speed”, even though the actual page load time was the same.
Initially this required the Developer Tools or a Terminal invocation via the defaults utility, but then a small shareware economy of GUI utilities grew, promising to “accelerate” or “optimize” Safari render speeds, with packages like SafariSpeed. Turns out that ever since Safari 1.3, WebKit developers had deactivated this preference key, so the trick was rather useless. Interestingly enough, some people felt the speed up anyway.
So a preference that was designed for purely cosmetic purposes in the first place managed to, even in death, soothe the users’ need for speed.
Any placebo effects with software that you’ve experienced? Is there a place for these kind of tricks in user interface design, if only to get the would-be tweakers and l33t hax0rs to pipe down for a while?
via Surfin’ Safari