HOLY MOTHER OF CRAP!
Apple changes the rules again - they've literally reversed the scroll!!!
My internal jury is still out...
While it's feeling so deeply wrong to change such rules, it actually feels somewhat right (if you forget the paradigm)
Apple's made a focus on bringing some Natural User Interface (NUI) concepts from their iOS platform to this version of Mac OS (see their 'Back to The Mac' keynote); this concept of direct manipulation is one of them. In the system preferences, the feature's even labeled as 'Natural' scrolling mode.
The introductory video on first boot-up of the OS introduces this feature via a short video to prepare users for the change, which indicates that Apple understands what a challenge it'll be for users to understand and get used to. In my opinion, it's one of the more radical/risky aspects of this new OS, in that it attempts to reverse a well-understood, long-standing paradigm for interfacing with the operating system.
Notice the default Lion UI also hides the scroll indicator until you're actually scrolling a la iOS, which is probably intended to reinforce the idea of 'pushing' the document up and down vs. focusing on the scrollbar. However, both of those (scrolling direction and scrollbar visibility) can be changed back to 'normal' mode in the system preferences.
I installed a system hack on Snow Leopard to simulate the new type of scrolling on my work computer before Lion was released and found it very tricky to change my thinking from being scrollbar-centric to document-centric even though I own an iPad and iPod Touch, but I eventually got (mostly) used to it on the desktop. Anyone who borrowed my machine when we were collaborating on projects, though, thought it was some kind of prank :)
So sometimes the onus is on the end users. Interesting to see how this turns out - but the bottom line is that we will adapt. Design power peeps.
I assure you that it takes an hour or two to mostly get used to it, and within two or three days you'll find the old scroll (really reversed, the new one makes more sense from mutitouch trackpads and such) to be unusable...
On Sat, Jul 23, 2011 at 10:17 AM, Ari Kolbeinsson wrote:
> I assure you that it takes an hour or two to mostly get used to it, and
> within two or three days you'll find the old scroll (really reversed, the
> new one makes more sense from mutitouch trackpads and such) to be
Question to anyone who's upgraded --
Is this applied across the board to all apps including non-native ones?
e.g. Adobe Air-based, Java-based (NetBeans, jEdit), X11, etc.
AFAIK yes - I tried it on Flex Web and Flex AIR apps and the scrolling API sends the 'correct' new signal to the Flash-based apps so a 2-finger swipe up now scrolls up instead of down, even in Flash stuff. The Flash scrollbars are still there and they still function as expected.
I also noticed it in Eclipse/Flash Builder, which is Java, that all the scrollbars are gone and the swipe matches Lion's preferences.
Took me a while - but I'm starting to get it. Same with the wife. She had the same (and uninfluenced) conclusion I had - that the iOS and normal OS worlds are converging and it actually makes a bit more sense, after using the iPhone and iPad for a while, that the gestures should be the same on all devices - even if it takes a few days of getting used to.
Weird though - that Apple has us drinking all this Kool-Aid where they can say - hey - you're holding it wrong, swipe it like this etc. and all we, like sheep, go - oh, okay. :-P I wish I was Apple too.
Brandon E. B. Wardbrandonebward@gmail.com
UI • UX • Ix DesignFlex • Flash Development
The original Lisa and Mac vertical scroll arrows were at the top and bottom of the vertical scroll bar, and the up-pointing arrow moved the content down. I ran a user study in the early days of Lisa development that informed that design.
Most (but not all) study participants expected to position the mouse near the top of the window to bring the content hidden above the top of the window into view. One reason was that they were looking at the top of the window at the time. Another reason was that they were more likely, as their next action, to select content in the upper half of the window than in the lower half. Consequently, we made the upper member of the arrow pair move the content down. With apologies to computer architects, I'll call the majority whose expectations were met by this decision the "top-endians".
The study also examined the question of which way the arrowheads should point. Half the participants thought the upper arrow should point down, the way the content was moving. Half thought it should point up, the direction from which the content was coming. If the latter is surprising to you, consider that a wind blowing air from north to south is called "northerly" in English, and that a standard PRNDL floor shift makes the driver push the stick forward to go into Reverse and backward to go into Drive. If you don't accept those analogies, note that the "elevator" in the scroll bar moves upward when the user presses the up-pointing arrow to scroll content down.
Most of the product team wanted the arrows to point the way the content moved, and to point away from each other. The issue was escalated to Trip Hawkins, the VP of Lisa Product Marketing. After hearing the study results, Trip offered to go with the usability study on arrow locations if I'd be willing (which I was) to make the arrows point away from each other because it "looked right".
Years later, Apple gave users the option of placing the scroll bars together at the bottom of the vertical scroll bar (as in the NeXT OS) or apart. The default became "together". When arranged in this way, users no longer had as much reason to expect the upper arrow to scroll content down. Top-endianism lost its logic.
As scroll wheels proliferated, the direction of the user's finger on the top of the wheel matched the direction of the scroll arrow and the "elevator", not the direction of content motion. I found this counterintuitive at first. I got used to it when I began to perceive the phenomenon as the bottom of the wheel pushing the content the other way. PRNDL again.
Some iOS apps that provide remote login to desktop computers originally required the user to push the content up to make it move down. I found adapting to that gesture hard. With a touch screen, the intuitive gesture to scroll content down is to drag any part of the content downward. (Of course, it is also intuitive for a drag to select a content range and for a drag to move an object without scrolling the surroundings--but that's another discussion.) Top-endians should be happy with the Lion defaults because the scroll gesture can begin anywhere, even the top of the window, which is where they are looking. Bottom-endians should be happy because there are no arrowheads to challenge their intuition.
I think people who use scroll wheels a lot may take a little longer to adapt to the Lion default than those who generally use scroll bars and function keys (page up, etc.).
On Jul 23, 2011, at 5:18 AM, Queen Catherine wrote:
> HOLY MOTHER OF CRAP!
> Apple changes the rules again - they've literally reversed the scroll!!!
> My internal jury is still out...
> While it's feeling so deeply wrong to change such rules, it actually feels somewhat right (if you forget the paradigm)
Nice post Larry. The problem with the approach is that it ignores the keyboard (as articulated). I am assuming this is a cross-platform issue...please correct me if I am wrong and thanks again for the explanation of due diligence.
On 25/7/11 9:12 AM, "Larry Tesler" wrote:
> If the latter is surprising to you,
> consider that a wind blowing air from north to south is called "northerly" in
> English, and that a standard PRNDL floor shift makes the driver push the
> stick forward to go into Reverse and backward to go into Drive. If you don't
> accept those analogies, note that the "elevator" in the scroll bar moves
> upward when the user presses the up-pointing arrow to scroll content down.
Also, when driving a car and you want to shift the object in frontal view to
the left (eg. a whomping great tree), you pull the steering wheel to the
right (yeah, ok, the top of the steering wheel).
Of course, the user's mental model is most likely that they are influencing
the position and direction of their vehicle, not rearranging the whole
universe around their frame of reference =P
Meanwhile, in many 3D games it's common to find a "reverse mouse direction"
option. I play in the mode where I point where I want to face, which means I
move the mouse to the left to look left ... and see the whole world shift to
the right. So, every time I go to google street view I end up getting
confused, because google street view wants me to imagine grabbing the
universe and shoving it around with my godly fist.
As scroll wheels proliferated, the direction of the user's finger on the top of the wheel matched the direction of the scroll arrow and the "elevator", not the direction of content motion. I found this counterintuitive at first. I got used to it when I began to perceive the phenomenon as the bottom of the wheel pushing the content the other way.
I'm glad I'm not the only one who came up with this analogy. In explaining to my wife the difference between the old and new way, I explained that in the old way it was as if the piece of paper was underneath the wheel, and thus when you roll the wheel up (away from you), it pushes the paper downward (toward you). In the new metaphor, you're now interacting directly with the piece of paper. This isn't necessarily entirely intuitive with a wheel, but it seems to make much more sense when using a touchpad-style interface, which (since the Magic Mouse now has a touch surface) is the only thing Apple sells.
It also, of course, lines up with iOS (and a theoretical future Mac with a touchscreen), and there are a heck of a lot more iOS devices in use than Macs.
It's funny, because when Microsoft a while ago ran a TV commercial demonstrating the use of a touch UI on Win7, it looked absolutely ridiculous to have the user moving her hand downward (dragging the scroll bar with her finger) while the content moved in the opposite direction. At that point I realized that the hardware UI metaphor we've been using since the advent of the scroll wheel was out of synch, and presumably would eventually change. Which it already has for Mac users.
Surprisingly, it's only taken me a couple of days to mostly adjust to the change (and I'm not a heavy iOS user, either). I expect within a week I will be doing it without any false starts in the opposite direction. Of course, I also still use inverted controls in FPS games, but that's relying on muscle memory dating back to the N64.
The piece of paper analogy is deeper than the use of a scroll wheel. Many, many years ago, I was involved in just this discussion concerning scrolling on DEC terminals. The way I described the two approaches was: Imagine that all the content is on a large sheet of paper. You can imagine "scrolling" (the word is actually biased here) in two ways: Either your have a fixed window that is looking at a large piece of paper, and you move the paper around behind the window to make different parts visible; or you have the same piece of paper tacked down, and a viewer - think of a magnifying glass - that you can move around to different places to see them. The first gives you "down means down"; the second gives you "up moves content down".
A scroll bar is a direct representation of the "fixed paper/moving viewpoint" model: The bar is the forever-fixed piece of paper, the scroller your viewer. If you have that, it drags the rest of your interface along with it. On the other hand, a direct-manipulation finger interface has you "touching" - and moving - the "paper", not the (on small devices) invisible frame, so the opposite model becomes almost forced. If, rather than using touch, you used the motion sensor on the device to implement scrolling - move device up to look at stuff further up the imaginary paper - you'd come to a different conclusion about what's "natural". (Not that I'm recommending that as an interface!)
This reminds me of the classic UI question: Given a pair of cold and hot water handles, to turn them both on, would you expect to (a) turn them both clockwise; (b) turn them both toward the tap? If you actually ask people, the context is determinative: If the handles are circularly symmetric (e.g., cross shaped), people expect to turn both clockwise; if they are not (single paddle), they expect to turn them toward each other.
Its now up to the users to get used to the new scrolling system. At first, it takes a lot of time getting used to, but after using the various apple mobile devices, scrolling the new way becomes second nature. It was a very risky move for Apple to make the new scrolling interface the norm, but I think it will pay off. They always seem to know that consumers are willing to bend to their rules.
Paul - http://www.connetu.com
I think that whatever changes Apple make to the UI are not taken lightly, as they are bound to have a team dedicated to researching and testing these changes before rolling them out. That being said, it still can take some time to adapt.