NN/g Home AskTog Interaction Design Section Apple's Flatland Aesthetic, Part 1
AskTog, February 2009
Appleland is becoming progressively flatter and, at the same time, less usable.
Apple has released a series of revolutionary products over the last several years, from System X to the iPhone. All represent Herculean software efforts. With such marked changes, one can expect that early releases will tend toward the primitive. Over time, users can expect missing functionality to fill in. For the most part, this has occurred and will continue to do so, with even highly-sophisticated features appearing, such as copy and paste.
Bruce Tognazzini was hired at Apple by Steve Jobs and Jef Raskin in 1978, where he remained for 14 years, founding the Apple Human Interface Group. He remains a major Apple fan, which is why, when they're doing wrong, he feels compelled to talk about it.
Filling in obvious features, however, is only one aspect of software evolution. Equally important is keeping up with the users. The beginner today will be the expert of tomorrow. The user with 200 photos today will be the user with 2000 a year from now. The user with 10 songs today will be the user with 100 songs six months from now. The user with one or two extra apps on the iPhone will be the user with 100 apps three months from now.
The great software applications usually didn’t start out that way. 20 years ago, there was a simple application on the Mac for doing basic edits on photos. It was called Photoshop. Today, Photoshop is a powerhouse of sophistication, capable of working miracles in the hands of a professional. Adobe has been in lock-step with their users, increasing Photoshop’s sophistication even as their users increased in theirs.
Apple, on the other hand, nowadays comes out with an initial, visually-simple model, and then just sticks with it as their users quickly outgrow it, resulting in their software appearing, over time, to lose power, instead of gaining it. For several years in the Jobsian Regime, I put down this perpetual primitivism as just a lack of resources. Now, I think it’s more. I think it’s a naive, misplaced belief that flatter is better.
The new Apple seems to subscribe to the belief that visual simplicity equals actual simplicity. This proposition indeed holds true on day-one of a person’s adventures in computing. It may even hold true on day 90. It does not hold true in years one and three and ten, when the user is struggling to corral thousands of documents with the same tool that was visually-optimized for 20.
Properly-designed interfaces scale, so that they support the new user as well as the expert. One of the beauties of Photoshop is that, even after all these years, with all its increased complexity, it still is able to be a simple application for doing basic edits on photos. (A new user can become productive in Photoshop in 10 minutes, even if it takes another 10 years to learn everything.)
So how do you present a new user with an object that is not overwhelming, while at the same time offering advanced users the ability to handle greatly-scaled tasks? In some case, you can design a single object that scales from the simple case to the highly complex. The Folder is such an object. A folder with 20 documents is visually simple, but so is a folder with 20 folders, even though each of those 20 folders, in turn, may contain another 20 folders, each with 20 documents, for 8000 documents in all.
The second strategy is to migrate users to more complex objects by offering an object of utility matched to the needs of new users, then replacing that object with one of more complexity later on.
For example, the iPhone user may start out using the supplied Weather app with its useful, but primitive level of information"it's gonna rain today"but shift to accuWeather in the midst of their first thunderstorm (such as I'm experiencing as I’m writing this). These users already know it's raining outside; they want to look at a radar map to see how much worse it's going to get and how long before it's over.
Having a single, perfecly-scalable object is the ideal solution, since it means the user doesn't have to start over, but having multiple choices of objects may be necessary if the evolving complexity is sufficiently advanced.
Let’s look at several examples of where Apple, on the Macintosh, is doing neither, but is, instead, just sticking with a beginner object, even when that object has long since become overwhelmed. In each case, you will see that the beginner object is flathierarchy freewhich is both what makes it so approachable for the new user and so utterly useless for the advanced user.
Apple’s flatness mania first became apparent with the Macintosh Dock. While an effective sales tool, it rapidly becomes a hinderance as users progress. The Dock can handle perhaps 12 or 15 applications sort of well.
As an experienced user, I can't use the Dock to hold my running applications and my 461 current favorite applications, folders, and documents because I don’t have a ten-foot wide screen. Instead, I used two separate shareware applications to cover both of the Dock's two main functions. One, called ASM, shows me all my running applications, including their names, in an alphabetical list, without taking up any precious desktop space to do so, since it displays them from a pull-down menu in the menu bar, just as System 9 used to do.
I gain single-click access to all 461 of my favorites by using another small shareware program called DragThing. Whereas the Dock is flat; DragThing is not. I’ve created 19 different tabs in DragThing that encircle my desktop. Hovering over any one of them opens its drawer, displaying all its files and folders. For example, my Database tab shows all my databases for immediate access. (DragThing offers the added bonus of letting you restore the Trash Can to the desktop, where it belongs.)
The Dock does an excellent job of representing to potential customers the open, friendly face of an Apple computer, however, it doesn’t even begin to scale. This is a situation where Apple needs to offer a second, compact, hierarchical object that is still Mac-like, but scales beyond the needs of the beginner.
Safari's add-a-bookmark opens up all the bookmark folders at once, giving the user a single, linear list. The new user can then plunk down a new bookmark among the eight or ten existing ones without having to worry about opening folders to do so. Nice. It’s not much of a solution, though, for that same user a year later, when he or she has hit 200 bookmarks and the list is seven or eight screenfulls tall.
What should Apple do? Is this a case where it becomes time for users to switch to a more complex mechanism? No. The fundamental approach is backwards. Apple is already providing an excellent mechanism for selecting a bookmark. Why not use that same mechanism for storing new ones?
To add a bookmark, the user should just pull down the regular bookmarks, navigate in the usual way to the subfolder desired, then release on an “add bookmark here” item found within each subfolder.
Comedian Steven Wright once announced that a friend had given him a map of the United States, “actual size.” When you’ve already given your users an actual-size map of something, be it bookmarks or anything else, it is usually a mistake to come up with some alternate way to perform the same navigation.
(Fortunately, there is a work-around to the bookmarks problem: Open a second window and open up all bookmarks within it. Then, drag the Favicon [the icon next to the URL] to where you would like it to be.)
iPhoto aggregates incoming photos in objects called, “Events.” A typical event might hold all the photos you took at the baseball game last weekend. By adding keywords to the photos, they can be later drawn up into albums reflecting, for example, all ball game photos you’ve ever taken or all photos featuring your favorite player at bat. iPhoto allows you to view just under 30 events at once, all laid out with giant thumbnails, so, for the first few months, the mechanism works quite well.
I currently have 522 events holding more than 13,000 photographs total. It takes 19 clicks in the scroll bar to view all 20 screenfulls of them and an average of 38 clicks to find the one I’m looking for, since I usually miss it the first couple of times around. Apple provides no alternate view other than giant thumbnails, and, in keeping with Flatland, no way of dropping groups of events into folders.
Keywording lets me overcome what would be a fatal design flaw, since I can zero in on a particular event by typing in keywords I know are associated with it, but this scheme works in spite of Apple’s best efforts, not because of it.
Keywords have also been kept linear, with iPhoto just producing a gigantic blob of text displaying all the keywords in a matrix, and, by "all," I mean about 20% of them. (The rest remain permanently hidden from view, since the blob isn't big enough to show them.)
Fortunately, for those people who have outgrown iPhoto’s native capabilities, Shareware has again come to the rescue in the form of Keyword Manager from Bullstorm. This add-on is not only greatly superior to the Flatland keywording system within iPhoto, it beats the keywording system within Aperture, Apple’s pro version of iPhoto.
With Keyword Manager, users can nest their keywords as deeply as they wish, and choosing a keyword deep in the hierarchy will automatically also keyword the photo with every keyword in the hierarchy above it. For example, if I add the keyword, Aircraft Carrier, it will add Warship, Ship, and Vessel. Later, searching for any of these will find the photo.
When I open Keyword Manager, I see 14 categories. When I open Apple's presentation, intended to be visually-simpler, I see more than 100 keywords. I actually have over 1200 keywords, and I can get to any one within three clicks with Keyword Manager. With Apple's scheme, I can't get to them at all unless the keyword I'm looking for is in the 100 iPhoto will show me. The rest are permanently hidden.
When an aesthetic gets this much in the way, it's time for it to go. The fact that it's hung on this long is rather amazing, since Apple tends not to cling to bad ideas. I suspect this bad idea has been hanging out in Apple's collective subconcious for several years now, unquestioned and, with probably a great deal of effort, somehow not noticed, although how you don't notice for eight years that 90+% of any serious user's keywords are just simply inaccessible is beyond me.
Apple needs to take a fresh look at all of their products across the board, specifically looking for where old decisions favoring new users are now dragging those same users down. Of course it's a good idea to avoid complexity, including hierarchies, where possible, but some tasks are inherently complex. Go for visual and behavioral simplicity where it works, but be prepared to back off.
Next: Part 2: Flatland & the iPhone/iPod Touch, making life very difficult for users and on the verge of costing Apple’s independent developers millions of dollars.
Join my intensive (and fun!) lecture/ workshop course. Sign up now!
Interaction Design course: Go from zero to interaction designer in just three days.
User Experience Conference Website There's more than my course at an NN/g conference. You'll find a breadth of other specialized courses and networking opportunities that will put you and your company at the leading edge of the design curve.
Have a comment about this article? Send a message to Tog.
Previous AskTog Columns >
Contact Us: Bruce Tognazzini
Copyright Bruce Tognazzini. All Rights Reserved