The point about angle of escape was really interesting. However, my google fu failed me. The only relevant link mentioning angle of escape leads back to this post. Is there someone who can give me more information about this? (Or a more technical term, if the author uses the term colloquially).
Once again, UI design never fails to fascinate me. The first concrete reason I ever came across that explained the superiority of the Mac user interface was the application of Fitt's law to the menu items. This looks like another solid point in favor of the Mac.
Look closely at the image he includes. He draws a line diagonally through each of the respective Apple and Ubuntu menu headings being discussed, in each case from the upper, outer corner of the menu item to the opposite, lower, inner corner. This line forms an angle with the neighboring vertical edge of the corresponding menu (which is more or less also the edge of the display area/screen). That angle is what he is describing.
In Apple's case, the angle is wider, making it -- in his description -- easier to move the pointer from the menu header, after triggering the menu, onto the menu items. In Ubuntu's case, the angle is narrower. Further, in the Ubuntu case, since the menu text is justified against the far side of the menu, the pointer has to travel further to hover over it and, for the top few items, that travel cannot be a straight line because a straight line would cross the neighboring menu header and so trigger the neighboring menu to drop down. At least, that's how I interpret his argument.
Looking further at his image, I notice that he is including some horizontal space between the Apple menu header and the screen edge as being part of the Apple menu; however, that area carries a different coloration. Not having a Mac at hand, I'm unable to test whether that area also functions are part of the Apple menu header. If not, then he's calculated the Apple "angle of escape" incorrectly.
Also, on my installation of Ubuntu 11.04, running Gnome Classic (aka 2.whatever), in the menu under discussion, one does not have to hover over the text in order to activate a menu entry. In my view, this makes the "angle of escape" argument somewhat less important -- at least, under Gnome 2.x -- in that one can use menu items by moving the pointer in more of a "straight down" direction. Although, the inclination may still be to move the pointer over the menu item text, thus requiring the user to learn to avoid attempting such a straight line pointer movement for the first few menu items on this menu. I don't know whether Unity -- the Ubuntu desktop environment he is actually describing -- handles these menu entries and their activation differently (my Ubuntu machine's old graphics processor can't handle Unity -- to which I sort of say, yay!).
The image is correct, while only small space around the Apple logo highlights upon clicking, the clickable area for the Apple menu runs completely to the left edge of the screen. The Spotlight icon behaves similarly.
I don't think a proper term has ever been used before now. There was a lot of thought put into the angle of escape when accessing sub-menus in the pull down menus. See the answer to "Question 6" in the link below. If anybody know the proper term, Tog would, and he doesn't appear to know of one.
I’m curious why you seem to think that Fitt’s law doesn’t apply to touch interfaces. Sure, certain concepts can’t be translated 1:1 (like infinite sized targets at screen edges) but even those have somewhat similar counterparts in touch interfaces (screen edges are also special places with larger targets in touch interfaces).
There's no reference to this in "Tog on Interface" either, but he repeatedly references Fitt's law. I think that applies to the "escape angle" as well.
I think it follows from Fitts' Law or rather is its reverse:
If the cursor is outside an area, the bigger and closer this area is the easier it is to purposefully hit.
⇔
If the cursor is inside an area, the bigger this area is and the farther away its borders are, the harder it is to accidentally leave.
"The steering law in human–computer interaction and ergonomics is a predictive model of human movement that describes the time required to navigate, or steer, through a 2-dimensional tunnel." -- http://en.wikipedia.org/wiki/Steering_law
Every source in the wiki article is either not linked or an article at ACM. Pay wall is a pity. I wonder if ACM crawl wikipedia looking for articles hosted on other (free) sites and link them to the same ACM hosted article?
EDIT: I've now added all the real links I could find. I'll check back in a few weeks to see if they have all been removed.
I find the author's implication that users select menu text (as opposed to any part of the entire row) interesting. I always move my mouse until the desired item received highlight rather than aiming for text, so the Ubuntu power menu was never weird to me. I wonder how prevalent either behaviour is...
It wasn't until your comment that I understood what he was talking about. I just couldn't fathom the usage scenario he's describing -- I've never seen anyone use menus that way.
Once again, UI design never fails to fascinate me. The first concrete reason I ever came across that explained the superiority of the Mac user interface was the application of Fitt's law to the menu items. This looks like another solid point in favor of the Mac.