I feel kind of sad for the guy every time he gets a public soap box from which to yell.
In one way, he's actually correct; the computer UI could have gone a lot of places since 1984, but it hasn't gone nearly as far, or nearly as efficiently in the right direction, as it could have.
And he's also partially right that the current iteration of the MacOS interface is a bit sloppy; it draws from several different design lineages, and contains an unfortunate amount of inconsistency--you have a mix of the OS9 era finder window with icons plus menubar interface, then you have some apps with the classic Photoshop-style floating toolbar, then you have the early-vintage OSX icon menus at the top of windows, plus the iTunes Drawer concept, plus the "appliance" style of more fixed interface with the brushed metal. Almost all good ideas, but no, neither invisible nor entirely consistent.
However, what theoretician Jef apparently fails to grasp is that it's an evolution, and a hugely extensible OS with 10 years of legacy following it around (Adobe isn't about to completely reenvision Photoshop) isn't the same thing as a car or a microwave. OSX, as it is now, is quite useable, and getting consistently better; the various different design lineages are slowly merging (Tiger previews are showing this progression) into something more consistent and efficient--we're evolving in the right direction, so to speak.
Theoretical types sometimes have great ideas, and their concepts can drive people in the right direction when applied properly. But in most cases they also fail to take into account the fact that we live in a real world where, in this case, a thousand different developers make ten thousand different apps, some with code bases more than a decade old. It is neither unified nor perfect, but in many cases it is the most efficient--in terms of development time and learning curve--way of doing things.
Go take a vacation or something, Mr. Raskin--maybe you'll feel better after you've stared at some trees for a while.
