I’m getting a divorce.

Today I officially announce my separation. It’s been a long two years, filled with triumphs, setbacks, and shouting myself hoarse. We’ve gotten an awful lot done together, and we’ve been as close as I’ve been angry (and about as often, too).

Maya and I are getting a divorce. We’ve come to a reasonably amiable agreement, and the separation should be a relatively quick and painless one. We’re working together on completing one last project or two, and after that, all hats are off. We’ll see how things fare, but we’re no longer relying on each other for personal or professional needs.

How did this happen? Well, the story demands some history first.

When I was boy of 11, I discovered Bryce through a magazine. Bryce seemed kind, open, powerful, easy to read, and really pretty, too. I had no idea at the time, but Bryce would profoundly shape my development.

Boy was Bryce great. Looking back, I’m almost chagrinned to realize that I was with him for so long, but he was one hell of a partner. Back when I was just a little kid, Bryce made me feel powerful. With Bryce, I created things that impressed my friends and parents, and there was a time when Bryce was the one I turned to when I was bored or had a couple minutes of free time. In fact, looking through my file logs, I see now how astonishingly productive Bryce made me. Sure, he was a bit thick at times, and I’ve certainly moved on, but Bryce was the best possible thing that could have happened to me at that age. Bryce sparked my interest in art and computers, and he fed me a steady diet of accomplishment that gladdened my parents. Bryce made creation a leisure activity.

For seven years, I was blissfully happy with Bryce, and we had so much fun together. But gradually, I became aware of Bryce’s limitations. Sometimes he would do things “the hard way” and make me work around it. Other times, I just got the feeling that as warm and happy as Bryce made me, I was capable of more than he had to offer. Eventually, I realized that I had to move on. It wasn’t easy, because I loved Bryce and was understandably apprehensive about leaving the stability of my comfort zone to enter the outside world. But Bryce had made me realize that my passions lay in 3D, and that he could never satisfy them alone.

I can’t remember how I found about Maya. Maybe it was a friend, maybe I heard about her through the grapevine, or maybe I knew about her for a while and just got up the courage to try her out once I had a reason. Whatever it was, we got together in February of 2005 and I was convinced that it would be a beautiful partnership. Poor naive me.

At the beginning, I admit that Maya well, kind of bowled me over. She was so experienced, and she could do so much, that I was a little intimidated. I made some tentative steps at doing things together, but it was very clear very soon that I was way, way out of my league with Maya. We talked things over, and I agreed to come back when I had some more experience.

I got the opportunity to acquire it when I went to college. In fact, there was a class available specifically for that purpose, and I hurled myself at it. Maya and I were together again. I learned about her personality, her quirks, her flaws, and all that she had to offer me. It was vast. She could do things I’d never even heard of, and she could do it well, too. The only problems were some communications issues we ran into, but I was confident that they would fade as we got spent more time together and got closer.

Again, Poor naive me. As I got closer to Maya and spent more time with her, I became more and more frustrated. Maya seemed to prefer math to art; she made me futz with numerical values and text fields and learn what I hated in high school, just to create art. There had to be a better way, but Maya didn’t know it.

Gradually, my dissatisfaction with Maya increased. So did my productivity and output, interestingly enough, but I always felt like I was fighting Maya to get things done, rather than working with her; I felt that I could be doing even more, even faster if only Maya would cooperate.

I eventually came to realize that Maya and I had fundamental personality incompatibilities. Maya likes absolute maximum choice and control, and she doesn’t like to teach people how to do things–she suffers no fools. I, on the other hand, appreciate a helping hand and gentle mentorship through new and difficult skills and ideas. Maya was spectacularly unsuited to deliver what I craved so badly. The friction increased.

Then Maya began to destroy my work. We would be in the middle of a project together, when she would scoop up a model I was working on and leave. I’d try to follow her, but she was too fast for me. When she came back, the model would be gone and she would wryly claim to have no knowledge of what happened.

I would yell and scream, but none of it ever made any impression on her. She would sit there and mock me.

Slowly, I began to dread talking to her. I avoided contact. I kept to myself. I stopped doing what I loved because my partner had broken our unspoken agreements and violated my trust.

So I left.

Maybe later I’ll talk about my future plans, but for now, I’m just glad to be rid of Maya.

Momentary solitary Mac OS X wish list

A friend once thusly described the primary difference between the way your choice of operating system affects how you think:

Windows is like a door painted shut. You say, “do I really need that door? No, it’s fine, I can do without it. Really, why do I even need that room at all?” While OS X is like the a beautiful door you look at and say, “Hmm, this is a great door–I love how smoothly it opens and closes–but I’d prefer a handle to this knob. It’d feel so much nicer–in fact, this knob is kind of uncomfortable. Yeah, the knob on this door is so frustrating! I hate this knob!”

The point being that Windows forces you to accept mediocrity, while OS X spurs you to constantly want more than you already have. Either way, it’s certainly true for me. When I’m in Windows I find myself accepting that some consistently forgotten preference (which of my screens is the primary one) will never be remembered and working around it. In OS X, I’m irritated by data lists that don’t have alternating backgounds and frustrated by buttons that don’t look identical to all other buttons.

Now, I’m a huge fan of Mac OS X, but sometimes just it pisses me off.

Let’s take, say, the font panel. Everybody loves the font panel; it’s easy to learn and totally consistent across application that make use of it–which is pretty much all of them. So what’s wrong with it?

Well, for one, it fails miserably at its primarily goal: allowing people to easily switch between fonts. With most pop-up font menus, such as the one found in Microsoft Word and practically every non-OS X word processor or text editor made in the last 15 years, here is the procedure:

1. Click on button to drop down menu of fonts.
   
2. Find a font you like.
   
3. Click on the font.
   

There’s nothing terribly difficult about that. Now, here’s the way you accomplish the same task with Apple’s font panel:

1. Click on the “fonts” button in toolbar or the buried “show fonts” menu item.
   
2. Move the font panel so it’s no longer in the way (somehow, it always opens somewhere where it obscures the text you’re trying to affect).
   
3. Click on a font to see what it actually looks like.
   
4. Click on the next font to see what it actually looks like.
   
5. Rinse and repeat until you find a suitable font after several minutes or searching, or give up in frustration and just use Helvetica or Times New Roman.
   
6. Close the font panel.
   

Notice any problems?
The primary, even crippling, disadvantage is that the fonts in OS X’s font panel don’t actually appear in the fonts themselves–in order to see what they really look like, you have to manually click on them one by one. And you get sick of clicking, it’s entirely too difficult to use the arrow keys. By default, font panels, like all inspectors, don’t retain focus when you click on something inside them–meaning that you can’t move between selections with the arrow keys, since the focus is still with the text you’re trying to wrangle a different font onto. You have to specifically click in the font panel’s tiny little titlebar–oh, oops, what’s that you say? You clicked and dragged? oh, no good, you can’t drag or it loses focus. You have to click and do nothing else but click, and only then will it retain focus.^[1] Then you can use the arrow keys.

The font panel seems to have been constructed with two seemingly contradictory assumptions in mind:

1. That people are going to be switching between multiple fonts often and rapidly, so it makes sense to put them into a window that stays open until explicitly closed (I’ll call these people Browsers).
   
2. That people know what fonts they’re going to use and don’t need to see what they actually look like (these people can be Zoomers).
   

Both of these assumptions seem incredibly wrong-headed to me–each one seems to actively annoy the other group! If you’re the kind of person who’s likely to ever change the current font (a Browser, most likely), you’re probably browsing for some unknown new one. This is a torturous process if you can’t see what your changes look like until after you’ve made them. The font panel is actively inhibitory to the browsing process these people go through when searching for a new font.

That leaves the Zoomers, people who know just what font they want. These people aren’t the ones who end up changing their fonts very often to begin with, so what’s the point of keeping the panel open for these people?

It seems to me that there’s a happy medium that could satisfy both the Browsers and the Zoomers. How about this: When the “fonts” button is clicked, the font panel opens right underneath it and disappears as soon as you click on one of the listed fonts–which are all shown as they will actually appear. BUT, this font panel retains its titlebar, so if you want to keep the panel open, you just have do drag it by its titlebar to somewhere else, and then you can click on as many fonts as you want without having it snap closed. That way it’s more like a pop-up panel, but it can become a full-fledged window if you deliberately tell it to.

^[1]
Note that I’m not trying to diss inspectors in general, because they’re not supposed to hog the focus. It’s just that I think making the font panel one was–along with many other design decisions made about it–a huge mistake.

Vista: An open-minded, Windows-using Mac fan’s perspective

Windows Vista is out, and there’s a lot to see. Much is pleasantly new, much is frustratingly old, but most of all, Microsoft’s hopes and fears are splayed out for all to see with this latest update to Windows. You see, Vista is Microsoft’s answer to the dual challenges posed by Apple and Google, and it shows; Windows Photo Gallery is an unabashed clone of Apple’s iPhoto, and offers to get you started with Windows Live are sprinkled juicily throughout the OS as an alternative to Google’s own web-based services. For what it’s worth, Windows Photo Gallery and its ilk are pretty good. Windows Live is not.

The most immediately obvious change in Vista is that it’s, well, no longer ugly; in fact, Vista is quite appealing. The new hardware-accelerated transparency and glass effects get the most press, but I’m most impressed by the use of subtle gradients, appealing high-resolution icons, and the general polish of the interface widgets. Buttons pulse and glow, windows fade in and out at the presence of the mouse, and work areas smoothly cross-fade into one another when you change categories. The overall effect is quite dramatic compared to the drab utilitarian look of Windows XP’s “Classic” look–itself a throwback to Windows NT–or the clownishness of XP’s standard plasticky bright blue theme. It’s obvious that a lot of graphic designers worked on Vista.

Pretty huge icons

Alas, the same cannot be said of interface designers. While Vista is certainly glamorously pretty, it’s still Windows under the hood. It still relies on wizards and cascading modal dialog boxes, still blurs the lines between applications and windows, still opens two instances of an application if you mistakenly double-click on it, still includes close buttons in every window’s titlebar regardless of how many other ways there are to get rid of it, and still jabbers to you with little balloons about anything that happens, ever.

As chatty as always

If the above traits or behaviors annoy you, you’re probably not a happy Windows user (if you’re even one at all), and Vista won’t do much to make you one. Like a fresh stab at a failing marriage, for these people Vista will be terribly appealing at first, but it hasn’t shed a single bad habit since last time. Though Vista dazzles you with lovingly crafted prettiness, it doesn’t hesitate to vomit modal dialog boxes all over the screen, ruining the effect. You can almost hear the graphic designers sobbing into their Wacom tablets every time you go to the Control panel to change something and wind up with three or four tiny cascading windows, transparently blending into each other and producing blurry visual mud.

Blurry visual mud

In many ways, Vista’s beauty displays Microsoft’s desperation. Vista almost seems too pretty; for example, in the bottom-right corner of new Explorer windows sits a strange cure-for-breast-cancer-inspired ribbon that shines and shimmers for a second for absolutely no reason. Why is this little thing there? What is its purpose? Who knows? It’s not part of Microsoft’s branding efforts, and it adds nothing whatsoever except perhaps to balance out the visual weighting of the bottom of the window (an icon always sits opposite it on the left side), but this ribbon thingy is pointless. Adding an appealing gradient to a flat-shaded title bar or a huge descriptive icon to a window is one thing; going around inserting random ribbons willy-nilly speaks of a lack of purpose and focus in the design.

Random ribbon

It’s almost as if Microsoft was trying to emulate Apple’s sense of beauty, all the while oblivious to the fact that so much of it comes from the spareness and simplicity of the design that Windows has always lacked. Microsoft’s ethos has always been to add on more features; Apple removes extraneous ones. Vista simply continues this trend; Microsoft layers on hardware-accelerated transparency that irks business users who must purchase more expensive computers to use it, and gamers whose games it slows down in a match of one-upmanship with Apple, trying to prove that it can make sexy things too.

Windows Branding Authority Enhanced
Microsoft liberally continues its extensive branding efforts in Vista by splashing the word “Windows” over everything it can get its hands on, winding up with “Windows Calendar,” “Windows Photo Gallery,” “Windows Mobility Center,” and, more awkwardly, “Windows Live OneCare.” This is nothing new; in 1996, the buzzword was “active,” producing nuggets such as “Active Directory,” “Active Desktop,” “and “Active Movie.” In 98, “Direct” was king of the hill, and “DirectX” and “Direct3D” reigned supreme. “MSN” wheezed its way through the early 2000s before being finally slain by “Windows” in the last couple of years in the grueling battle for the mindshare of a captive audience. One wonders what the next buzzword will be, or if we will simply have to live with those two extra syllables in the form of the word “Windows” appended to everything that comes out of Redmond in the near future. It’s already playing hell with alphebatized lists.

Menu bar, schmenu bar
Back in the present, Microsoft is trying to kill the menu bar. Arguments can be made that the menu is dated and obsolete, that visual metaphors are better at communicating related concepts and binding them to the content they affect, and a legitimate usability discussion on this subject exists. Apple, for example, has pioneered menu-less desktop modes with Dashboard and soon-to-be Time Machine. Apple has not, however, pioneered menu-less applications. Why? Because most applications need their menu bars! Especially complex ones, such as, for example, Windows Explorer, whose menu bar is strangely nowhere in sight when you first fire it up. Instead, you get a glassy bar with drop-down menu buttons labeled “Organize,” and “Views.” Depending on which folder you happen to be exploring and what happens to be selected, other buttons can appear as well, such as Explore, e-mail, Share, and Burn. The Organize menu button is a mishmosh of some commands typically found in old-school File, Edit, and Window menus, and the Views menu button lets you switch between six window views, two of which–List and Details– are almost identical.

The usability problems with this setup are immense; the easiest way to discover this is by revealing the hidden menu bar itself and seeing how much functionality had to be left out in order to accommodate those visually pleasing context-sensitive buttons. Scads of commands such as Map Network Drive, Copy to Folder, Invert Selection, and Create Shortcut are all hidden and generally inaccessible unless you know how to reveal the menus they reside in, or happen to have memorized their keyboard shortcut when the menus were visible.

However, in a bizarre twist of fate, those hidden menus themselves are context-sensitive, dynamically gaining and losing items depending on what’s selected–but get this: they can still contain grayed-out menu items! Isn’t the whole point of context-sensitivity that only what’s relevant is shown at a given time? Furthermore, what’s the point of having two interface conventions for the same tasks that are both context-sensitive? Shouldn’t the point of offering menus in addition to context-sensitive task buttons be to provide solid footing and comforting familiarity if you don’t like the constantly changing buttons? Making the menus themselves context-sensitive absolutely defeats this approach, and provides two redundant methods of doing the same thing with minimal differences in accessibility and efficiency.

Sometimes, toolbar buttons themselves are sometimes exceptionally menu-like. In the case of Windows Media Player, a series of glassy blue buttons with drop-down menu buttons replaces the menu bar. Clicking on the button itself changes the mode of the player, or drops down a menu if you click on the button of the current mode. If you want a drop-down menu for one of the other buttons, you have to click on the teensy-weensy arrow at the bottom of the button. The menus that drop down when you do one of these things are functionally identical to a real menu bar in that their commands are not context-sensitive like those of Windows Explorer are.

Sound confusing? It is, primarily since this new system duplicates about 75% of what goes into traditional menus, but with a pointless new visual style that ignores many of the improvements made to menus over the years, such as the ability to slide the mouse horizontally over to reveal a different menu. The end result is a strange mishmosh, a blending of toolbars, menu bars, and mode switchers, winding up with a bizarre hybrid that does all three in unfamiliar, unpredictable ways. As with Windows Explorer, you can opt to display the classic menus, which duplicates the functionality offered by the blue buttons. However, Windows Media Player’s menus are not context-sensitive like Explorer’s are!

WMP’s bizarre menu-button-thingies

This seemingly random inconsistency is maddening to me, and it calls into question just what kind of usability testing goes on at Microsoft and who’s doing it. Menus are time-tested for their efficiency and consistency. At the very least, all computer users in the last 12 to 15 years have used them extensively, building up vast unconscious reserves of knowledge about their workings and use. Overcoming this brick wall of experience would be difficult even with a superior alternative, and Microsoft hasn’t found one. Reinventing the wheel is bad enough. Trying to reinvent the rocket ship by replacing it with a wheel is quite another thing.

Okay, so Windows can’t match Mac OS X in consistency and interface efficiency. But it can try to compete on features and applications–especially some of the ones that have traditionally made Macs special. Windows Photo Gallery, DVD Maker and Movie Maker are all obvious shots at iLife, and the Windows Search is decidedly Spotlight-esque. But these are all defensive measures; they’re Microsoft pleading, “Aww, c’mon, I can make movies too! Honest! And look! I can find your files as fast as that Mac tramp can!” Some of these efforts do a good job of patching obvious holes in Windows’ capabilities; some do not.

Windows Search is an example of one that tries and almost succeeds. For everyday searches, it’s plenty fast–at least as fast as Spotlight, if not faster. But it’s nowhere near as comprehensive. Example: I create an RTF document on my desktop named Fungus.rtf and type a sentence inside it that contains the word “jello.” Searching for “jello” fails to find the file, so I search for “fungus”; that reveals it. I again search for “jello” and the file shows up; Windows Search only finds the file after it’s previously been found by Windows Search before–how useful is that bound to be? Indexing the contents of files is still iffy for Windows Search. Contrast this with Spotlight which uses kernel hooks to index files as soon as they’re created, in addition to several other advantages. In particular, the inability to search inside PDF documents is a real heartbreaker for me.

It’s Mac-y
Vista also includes some interface features that are clearly Mac-like in their aesthetic, such as Flip 3D. Flip 3D is a shiny new window management effect that shows 3D representations of all currently open windows (as well as the desktop) floating in mid-air above the desktop background, offset at around a 45 degree angle so you can see the windows behind the ones in front of them. You use the mouse wheel or arrow keys to move through the stack until you find the one you want by bringing it to the front and hitting return, clicking, or releasing the Windows key (depending on how you invoked it to begin with). In addition, you can also zoom right to the window you want by clicking on it.

Flip 3D is probably supposed to be Microsoft’s take on Apple’s Exposé–at least visually (“Look, my windows are flying around!”), if not in usability. Exposé is an effect that temporarily scales down currently open windows so they all fit on the screen, allowing you to click on the one you want. Unlike Exposé, Flip 3D displays all windows in a static line, making distant windows all but indistinguishable, and none of the windows are ever labeled, and the transparency only muddies things up even more. Basically, more windows makes the system difficult to use–a charge that can also be leveled at Exposé, as when a sufficient amount of windows are present, they’re all scaled down so far that their contents become similarly indistinguishable. But at least Exposé works well with under 12 windows–something that cannot be said of Flip 3D, and at least Exposé has the decency to label the window your mouse is hovering over; Flip 3D’s windows are never labeled at all.

In addition, the order of windows in Flip 3D isn’t even necessarily representative of their actual depths relative to each other. For example, just because a Firefox window is located directly in front of a Control Panel window in Flip 3D, that doesn’t mean that Firefox will be directly in front of the Control Panel outside of it. There may well be several intervening windows between them that were shown in completely different positions in Flip 3D. The non-representational nature of Flip 3D’s window display further hampers its usability.

In the end, though, it’s not these minor quirks that does in Flip 3D–Exposé has its share of them as well. Philosophically, Flip 3D is a disappointment because it’s really nothing new except a visual spruicing-up of an existing window organization system that’s still present: alt-tab. For those unfamiliar with it, when you hit alt-tab and hold down alt, thumbnails of the currently open windows appear on the screen, letting you choose which window you want by repeatedly hitting tab until the one you want is selected, and then releasing the alt key. If this sounds familiar to Flip 3D, that’s because it is; Flip 3D is, like alt-tab, a window-centric organization system–that is, it thinks in terms of windows and nothing else, such as applications.

On OS X, Exposé became a success because it was a window-centric organization system that complemented existing application-centric ones (The Dock and Apple-tab), rather than duplicating them. Flip 3D would have been a perfect opportunity for Microsoft to add an application switcher to complement the window-centric nature of the taskbar and alt-tab. Instead, it disappointingly offers yet another way to accomplish the same task without really rethinking the ideas of window management or window-centricity. Flip 3D is pretty, but I suspect most people will ignore it in favor of its faster and more discoverable colleague: the ever-present alt-tab of yesterwindows.

Windows Photo Gallery
Windows Photo Gallery, on the other hand, is a marvelous success. Though it’s transparently a Windows iPhoto clone (pun sadly intended), it’s really quite excellent. Unlike iPhoto, it’s sleek and light–it opens quickly, and it’s the default image viewer and editor–no more Windows Picture and Fax Viewer. Microsoft has finally come around to the merits of a library-based approach to organizing masses of files, but, unlike iPhoto, it’s pleasantly optional: you can do all your image viewing and editing without ever having to add a single photo to its library.

I like Windows Photo Gallery. It’s clean, intuitive, easy to use, and visually appealing. It has decent tools for editing and organizing photos, and I even prefer its tag system of metadata to iPhoto’s non-discoverable method for assigning keywords, as well as its “Tiles” view. Still, it’s no iPhoto in terms of raw power: it has fewer image correction options, no way to batch-change pictures, no smart albums, and can’t even save or export images in anything other than their native file format (you have to use the venerable MS Paint, amusingly/embarrasingly enough). As seems to be the trend, it’s menu-less, relying on context-sensitive drop-down buttons instead. Unlike Windows Explorer’s context-sensitive buttons, however, Windows Photo Gallery’s are always visible, just grayed out if they can’t be used. And there isn’t even an option to “show classic menus” like there is with Explorer. Sigh. Another blow to consistency.

Nevertheless, despite these limitations, I suspect that Windows Photo gallery will be just fine for the overwhelming majority of users who come into contact with it–which will be a proportion greater than those who use iPhoto on a Mac, I’d wager, due to the fact that it’s the default editor and viewer. Windows Photo Gallery has the makings of a success through and through–and Microsoft ought to be congratulated on it. Well done.

Gadgets/Widgets/Google Desktop
Some history is necessary before I talk about Microsoft’s Gadgets. It’s easy to yell, “They copied Apple!”, and “Apple stole the idea offa Konfabulator!”, and “It was Apple’s to begin with; they invented it in 1984 with Desk Accessories!” But this banter obfuscates a lot of the nuances of the issue. The history:

1984: Desk Accessories

The original Macintosh released in 1984 included little mini-applications called “Desk Accessories” that duplicated the functionality of desktop applications, but were lightweight and constantly available. Desk Accessories were created primarily to bypass the irritating hardware restriction of only being able to run one full desktop application at once; an unlimited number of Desk Accessories could run at the same time as your allotted one desktop application churned away at its 128K of RAM.

2003: Konfabulator

In 2003, Arlo Rose, Perry Clarke, and Ed Voas created a software environment dedicated to running visually appealing single-purpose mini-applications. Rose put a great deal of work into the artistic presentation of the Widgets, in particular achieving them a very “Apple” aesthetic by giving them rounded corners, photorealistic icons and glassy transparency.

2005: Dashboard

Apple released its implementation in 2005 to mixed praise and many charges of the theft of Konfabulator’s ideas, due probably to the degree to which Arlo Rose succeeded in duplicating the Apple aesthetic; when Apple itself released widgets that looked very similar, it pretty much appeared that digital plagiarism had occurred. Apple’s Widgets live in a summonable overlay that’s either invisible or taking up the whole screen; the difficulty in getting Widgets to seamlessly inhabit the desktop is frustrating to me.

2006: Google Desktop

Google Desktop tried their hand at the concept in 2006, but relegated its widgets to a customizable sidebar that pushed windows and icons over so they didn’t cover it up. This design worked very well due to the fact that since the sidebar is always active, information stored there is truly at-a-glance; all you have to do is look over a bit. The downside is that the screen can get extremely cramped if you don’t have a large or widescreen display.

2007: Windows Sidebar

The Windows sidebar is Microsoft’s take on mini-applications, and it looks about as identical to Google Desktop as Dashboard does to Konfabulator. As a clone of Google Desktop, it inherits all the advantages and disadvantages of Google’s product.

But just because the Sidebar is an obvious clone of its competitors’ products, that doesn’t mean that it’s a failure. On the contrary, it’s quite excellent. Both Dashboard and Google Desktop have limitations and downsides, and Gadgets take the good from both. While Gadgets are really nothing new, Microsoft implemented them in a way that I find to be more efficient than Apple’s and more integrated and widespread than Google’s. I’ve been itching for the Mac version of Google Desktop to imitate the Windows version, and, truth be told, I actually prefer Gadgets to Widgets. I like the idea of being able to keep working while Gadgets are open and visible; Microsoft’s approach offers more flexibility and doesn’t interrupt your workflow, and I approve of that.

The fact of the matter is that lightweight, single-purpose mini-applications are not new, and sidebars where you put things you want constantly visible are also not new. Heck, applications that combine both aren’t even new. But that doesn’t stop the Windows Sidebar from being really good.

The remaining kitchen sinks
Vista also ships with several other applications that appear to be heavily inspired by OS X, such as Windows Movie Maker, Windows Calendar, and Windows Mail. They’re all pretty good. They all have prominent limitations compared to OS X’s offerings, but also offer some advantages to compensate. However, none of them is sufficiently interesting to merit more than the acknowledgment that Microsoft finally recognizes the importance of bundling quality software with their operating system.

All in all, Microsoft’s response to Apple is well-thought-out, cleanly executed, and generally pretty good. Vista isn’t going to make any Mac users jump ship, but it will probably help retain some people who were envious of some of the Mac’s features and advantages. Dashboard, iLife, and Spotlight are no longer as tempting as they once were now that Vista has its own implementations of the basic ideas behind them.

And now for Google
On the other hand, Microsoft’s response to Google in the form of Windows Live is clumsy and uncoordinated. The main difficulty is that no one–Microsoft included–seems to know just what Windows Live actually is. Plenty of applications bundled with Vista have references or links to Windows Live services, but many of said services seem to be what could be perfectly viable desktop applications that have been clumsily pushed onto the internet to fill some marketing brochure. For example, Live’s Safety Center deals with securing and cleaning up Windows PCs and attempts to sell Windows Live OneCare, Microsoft’s pay-per-year security suite. There is no immediately obvious reason that the Safety Center could not be a desktop application as it involves the internet no more than any malware scanner does when it checks for updates to its virus list.

In many other cases, Live services seem to be Microsoft-branded carbon copies of pre-existing Google services. For example, on the Windows Live main page you will find services such as Search, Image Search, Video Search, Maps, Mobile, and Toolbar: add the word “Google” to the beginning of each, and that pretty much describes the innovation level present in Windows Live. The only problem in duplicating what the market leader does is that they do it better than you and they will win, simply because they’re bigger. This is a lesson Microsoft should have learned ages ago, because size is the company’s greatest advantage and one it’s used time and time again to push its technologies and products. Linux has for years and years been steadily trying to duplicate the functionality, features, and interface of Windows, not even making a dent in the battle for desktop market share, and it’s even free! How can Microsoft hope to do the same to Google with so awkward an effort?

In order to compete with a market leader, you have to offer something new. This is what Apple has been doing in response to Windows: Exposé, Dashboard, and Spotlight were all concepts that had never really been successfully commercialized on a large scale before Apple did it, and their inclusion in Mac OS X meaningfully differentiated it from Windows far more than the “look and feel” of OS X ever could to those who think in terms of tangible features.

Somehow, Microsoft fails to grasp this in its war for cyberspace supremacy against Google. It seems to think that merely duplicating and re-branding existing Google services and tying them to Windows will enable it to pull ahead. Microsoft is being proven wrong every step of the way, and the latest re-branding of its previous MSN efforts under the “Live” banner hasn’t helped anything. In fact, the Live home page itself even looks decidedly Googlesque with a search box front and center and offers for related services smattered around. It’s almost as if Microsoft is trying to woo people familiar with Google but looking for an alternative–an audience that I would guess is very small.

Windows Live homepage

Google homepage

Put simply, Microsoft cannot hope to beat Google at its own game because doing so would melt the company’s bedrock of profitability. Google is a company all about getting things off the desktop and onto the internet; Microsoft is a company all about dominating the desktop. This is why Microsoft cannot compete with Google on its own turf. Google’s success stems from its online search and advertising and its ability to compellingly webify bloated desktop applications, such as the desktop email client and office suite that Microsoft makes huge profits selling. To imitate Google would be to cannibalize its own business.

Why does Microsoft even have to beat Google, anyway? Microsoft’s profits stem primarily from licenses of Windows and Office and server products; Google isn’t particularly threatening to any of those–the people who perfer web applications weren’t buying Microsoft’s software suites anyway, and they wouldn’t buy one from Microsoft if it was bundled with the keys to Fort Knox. Why then is it so important to have an offering to match Google’s at all? It isn’t clear, and that’s why Windows Live is failing.

Windows Vista is an interesting beast. In some ways, it’s been totally revamped in pleasing and appropriate ways. In others, it’s just as clunky as ever, and more bloated to boot. It’s a mixture of beauty and shallowness, of great ideas and terrible ones. It’s better than ever and more of the same. I get the sense that Vista is more of a defensive release than an offensive one; it’s very Mac-like in some ways, and it tries to be very Googley in other ways. This is not to say that there isn’t anything innovative, because there’s lots of it. But most of it seems to be aimed at defeating Apple and Google, rather advancing what makes it Windows. It’s become directionless, stabbing out at any large tech company that achieves success, whether or not this threatens Microsoft or comes at its expense. That’s a shame, because it hurts both Microsoft and Windows.

But there’s no software for Macs!

“There’s no software for Macs!” is a common refrain in the Windows world as a reason to shun Macs. Along with “They’re more expensive!” and “They’re only good for artists and video editors and stuff!”, lack of software is part of a trifecta of shame routinely leveled at Apple by Microsoft-centric tech pundits. The problem is that two of the claims–”lack of software” and “only for artists”–are blatantly false, and the third, price, isn’t a problem if you’re the kind of person who appreciates the notion of paying a premium for premium tools. Regarding software, however, a more appropriate charge should be:

There’s waaaay more software for Windows than there is are for Macs!
Easily granted. Easily. But just because there’s more software out there for Windows–even a vast amount more–that doesn’t mean it’s better. On Windows, there are half a dozen pieces of software to accomplish literally any minor task (e.g. managing screensavers), and you can probably unearth 20 or 30 for big tasks (e.g.text editing). Many are good, many are shoddy; many are free, many are commercial; many were written by indie developers, and many were written by huge corporations. It’s going to sound bizarre, but having to slog through all that software can be a daunting task. This concept is well-explained by barry Schwartz in The Paradox of Choice, but in a nutshell, he explains that as consumption and purchasing choices skyrocket, people are more likely to be paralyzed with inaction. Who wants to spend all day deciding between 80 types of cracker? What about 40 kinds of stereo? 15 text editors? 18 antivirus packages? 22 PDF converters?

On the Mac side, there’s certainly less software, but it’s better. Even if you dispute this, you have to admit that there is a massive amount of crufty software for Windows, which makes it harder to ferret out the good stuff. In addition, there aren’t any good centralized online repositories of Windows software^[1], making it difficult to find quality examples of software written by independent developers. Locating good Windows software usually involves asking a knowledgeable friend, following links from sites you already trust, using Google and verifying with Wikipedia, or heading down to CompUSA or Best Buy, which are pretty much off-limits to independent developers who don’t have significant financial support.

On the Mac side, there are several avenues to go online window-shopping for software. VersionTracker, MacUpdate, and iusethis are three such big names, and Apple itself lists a huge amount of software on their site–which is accessible from the Apple Menu of any Mac:

These luxuries make it really easy to find excellent, user-reviewed software (in the case of VersionTracker, MacUpdate, and especially iusethis). Indie Mac developers typically do very well for themselves because it’s easy to find shining examples of the finest the independent developer community can offer online.

In addition, as Mac OS X is essentially a prettified BSD Unix, all Unix software will instantly work on any Mac. Linux and UNIX are related, so a lot of Linux software also runs on Macs. There are even active efforts at increasing compatability between OS X and Linux in the form of Fink and DarwinPorts. For any of this free open source Linux and UNIX software to run on Windows, it has to be partially rewritten, meaning that only the most popular will ever make it to Windows.

To prove the depth of Mac software, just take a look at the stuff Mac-heads actually use. You’ll probably find that while there are some commonalities (iTunes, Quicktime Player, Safari), most also use incredibly varied software from all over the map. For example, here are the primary pieces of software that I use, as well as the alternatives to them I could think up off the top of my head:

• OmniWeb (Web browsing. Alternatives: Safari, Firefox, Camino, Shiira, Opera, iCab, Seamonkey)

• Mail (Email. Alternatives: Mailsmith, Thunderbird, Eudora, Entourage, GNUMail, PowerMail, Seamonkey)

• MarsEdit (Blogging. Alternatives: Ecto, Journler, Bleezer)

• NetNewsWire (RSS reading. Alternatives: NewsFire, Newslife, Pulp Fiction, Vienna, Shrook, Seamonkey)

• Pages (Page layout/Decent word processing. Alternatives: MS Office, OpenOffice, NeoOffice, Mariner Write, Nisus Writer, WriteRoom)

• Maya (3D Modeling & animation. Alternatives: Blender, Wings 3D, Cheetah 3D, Cinema 4D, Modo)

• Photoshop CS2 (Image creation & editing. Alternatives: GIMP, GIMPshop, not really much)

• TextMate (Programmer-centric text editing. Alternatives: BBedit, TextWrangler, SubEthaEdit, Smultron)

• XCode and Interface Builder (Software design & GUI building. Alternatives: none)

• Parallels Desktop (Virtualization. Alternatives: VMware Fusion)

• QuickTime Player (Movie playback. Alternatives: VLC, Mplayer)

• iTunes (Music playback & organization. Alternatives: none)
  

Many are these products are specialized, and many are common, but almost all of them have a reasonable amount of alternatives. To Mac users out there, I encourage you to post your own abridged lists in the comments!

Aha, but I’ve got you now! Macs don’t have anywhere near the kind of software businesses need!
Ack, Achilles’ heel! It’s true, Macs simply don’t have much business software. It’s only recently that Macs got any decent POS software, and you can pretty much forget about specialized stuff like real estate or dentistry. It’s a problem, and one that Apple will have to overcome if Macs are ever to gain significant traction in major industries.

And what about games, you charlatan?
I have but a single response:

I should mention that I’ve been enjoying Half-Life 2 and Supreme Commander on my MacBook Pro for a while now. Boot camp is pretty much a revolution in Mac gaming. It used to be that Mac gamers suffered through poorly-done ports that were slower than their Windows counterparts and released a year or two later. Though the situation has hardly improved, it barely matters since we can now run Windows games at full speed! Ever since I got my MacBook Pro, I’ve opened up to the massive world of Windows games, and it’s great–especially since I can do it from my existing computer at no extra cost (besides that of buying Windows itself, of course). Truly, Boot Camp is one of the biggest advantages of today’s Macs. Due to Boot Camp and a thriving independent developer community, the software argument can be legitimately boiled down to:

There’s little dedicated Mac software for business or gaming though the ability to run Windows cuts down on the problem considerably.

So there.

^[1]
ZDnet requires registration and doesn’t have reviews; CNET sponsors those who can pay for advertising treatment and is much friendlier to commercial software as well having been accused of some dodgy practices; torrent sites don’t count.

Paul Thurrot gets it right

I often find myself at odds with Paul Thurrot’s opinions. As a Windows guy, it’s appropriate enough for him to prefer windows, but he occasionally says some things about Macs that I find baffling. Imagine my surprise, then, to find myself vigorously agreeing with him.

Paul Thurrot says that Macs are for more technical users. For years, he’s been repeating this line over and over again, and I’ve always dismissed it as pure fantasy. After all, everybody knows Macs are easier to use, right? The big selling point is the whole “it just works” aspect–clearly targeted toward new users and non-techies.

If only it were true. As bizarre as it seems, I believe that Paul Thurrot is right. I was reading this article of his in which he refutes the common truism that Mac and Linux users use their platforms because they want to, and Windows users use Windows because they have to. Thurrot says,

“I disagree.

But I think I see the point he’s trying to make.

…this isn’t “have to” vs. want. The truth is, for most people–like, 99.99 percent of the computing using public–a computer is a tool. What they “want” isn’t a particular OS. What they want is a solution to a problem, or an answer to a need. They want email. They want the Web. They want Office. Games. Digital photos and music.

Anyone who stays up at night worrying about OS platforms just isn’t part of the mainstream. That’s not good or bad, it’s just reality.”

I was all set to yell, “No, you’ve got it all wrong!” until I read those two paragraphs. The truth is, he’s right. Most people are astoundingly non-technical and have no idea how their computer works, or even if it is doing so correctly at all. For better or worse, 99% of computer users do indeed simply want to browse the internet, write papers and emails, play some games, and organize their pictures with a minimum of fuss, entirely regardless of what may be more efficient or appropriate. It’s just not interesting to them.

I can completely understand why someone would be bored by computers. I, for example, am entirely uninterested in cars. Detailed automotive specifications bore me–I don’t care how many Vs my engine has or how many horses one would need to outpace it. I want a vehicle that gets me to my destination with as little in the way as possible. I obviously care about safety, and a built-in CD player would be nice, but beyond that, my needs and interest level are pretty modest.

Most people are this way with computers, and Windows attempts to satisfy them. For example, Windows tries to phrase easy things in terms that non-technical humans can understand. For example, hovering the mouse over Internet Explorer’s icon always produces a tooltip saying something like “Browse the Internet”–a phrase universally comprehensible to everyone who has at least heard of the internet and knows it needs to be browsed. By contrast, Apple’s web browser Safari on OS X has no such labels hinting at its purpose–it’s simply assumed that Mac users know that Safari is a web browser. This assumption displays Apple’s knowledge of the technical bent of its users, and it also answers the otherwise puzzling question of why so many Windows switchers still use Firefox (or, *shudder* Internet Explorer) on their new Mac. There are also plenty of other nods to the differing technical abilities of the the platforms’ user bases sprinkled around. Here are a few:

Close window vs quit application
A convention many Windows users and switchers never learn when they have to use Macs is that of applications not quitting when their last window is closed. The principle behind this decision is that in an app that can open multiple documents or may need to operate invisibly, there is some utility in keeping it running when it has no windows open. For example, if you close the last open document in the Mac version of Microsoft Word, Word itself stays open, waiting for you to open a new one. If you close iTunes’ window, it continues to play music.

Most Windows users are entirely unaware of this behavior, and it’s obvious when a Windows user has been on a Mac because five to ten open applications with no windows will be sitting happily on the Dock waiting for instructions. If they are ever made aware of their “mistake”, your average Windows user will be baffled by the behavior. The Mac user will attempt to explain the “Quit” command and its accompanying keyboard shortcut, and how some applications may need to stay open in the background, but this is nonsense to most Windows users. OS X’s window management system actively takes getting used to to avoid cluttering the system with unused running applications–a departure from the friendlier system on Windows which assumes you’re done with an application when you close its last window. In short, the Mac system requires technical comprehension to understand and technical proficiency to use–both things that few Windows users are interested in possessing or acquiring.

Installing software
Mac users like to point out how easy it is to install software on a Mac. Download the disk image, double-click it to mount it, open the resulting volume, drag the application to your applications folder, then unmount the volume and throw away the disk image. What could be simpler?

Unfortunately, this task is not apparent if you’ve never done it before. For all the derision that Wizards garner in the Mac community, they are good for guiding those unfamiliar with the procedure for something. Installing Mac software requires that you know:
1) That you can drag things to your applications folder
2) That the disk image can be opened (i.e. don’t drag the disk image to your apps folder)
3) That the disk image is not itself the application (i.e. don’t drag the resulting volume to your apps folder)
4) That you shouldn’t run the application from the disk image (i.e. this is the one you do want to drag to your apps folder)
5) That can and should unmount the volume and throw away the disk image when the applications is in your applications folder.

That’s a lot of stuff to learn, and a lot of ways to fail if you don’t already know. Once you get the hang of it, the advantages of the Mac way become immediately apparent, but there’s no way to figure it out besided trial and error, being taught by another Mac user, or if the developer of the software has kindly put instructions in the disk image, like so:

However, this is only helpful if you’ve already opened up the disk image, and doesn’t tell you anything about what should be done once you’re done dragging it to the right place, nor does it provide an easy way to locate the Applications folder (though many do, helpfully enough).

Skype on OS X vs Skype on Windows^[1]
The Skype people understand the audiences of the platforms. As a cross-platform app, Skype has client programs that run on all three major OSs, but the most striking difference is between the Windows and Mac versions. Skype on Windows is very easy to figure out. All the buttons are prominently labeled, and common tasks are clearly spelled out in obvious places. My favorite example is the video chat interface. If a usable camera is present, a button entitled “Start My Video” appears. If the video is currently up and running, the button changes to “Stop My Video”. This is 100% obvious and extremely easy to figure out.

On the Mac version, however the video button is entirely unlabeled. Instead, it’s just a little symbol. Mac-heads will immediately recognize the symbol because it’s the same symbol used throughout the OS whenever video is present or possible. But for those unfamiliar with the convention, the button is cryptic. These people may well never figure out how to start up the video and wind up extremely frustrated.

The Skype team knows that Mac users will figure out that the tiny unlabeled symbol means “start video”. Does this mean that Mac users are smarter than Windows users? No, it simply means that they can count on Mac users being more technically inclined and more likely to experiment if the solution isn’t immediately obvious. That the Windows version of Skype involves more hand-holding speaks more to the technical ability of Windows users, not their intelligence.

The truth is, Mac users and Windows users really are different. Typical Windows users want their tasks done with a minimum of fuss, usually on the cheapest machine possible. Typical Mac users want and expect quality and luxury in their computers.^[2] Mac users understand the value of paying a premium for premium tools. They want systems and machines that “just work” not because they are unable to fix irritating technical problems themselves, but because they understand that doing so is a waste of their time. They constantly want more from their computers because they comprehend the value that smoothly functioning computers can add to their lives in the form of increased efficiency, productivity, and communication. They use their computer skills to create content for others to enjoy–in 1997, when Apple was floundering and suffocating under poor business decisions, around 64% of all websites were found to have been made on Macs.^[3]

Why then, do Macs have the reputation of being easier to use if their primary audience doesn’t need such ease of use? The answer is twofold: first of all, Mac users like things that are easy to use. They’re not fond of wrestling with their machines to get them to work; they’d rather be doing something fun or interesting or productive. It just so happens that Windows’s primary audience likes things that are easy to use as well. The second reason is that though Windows tries to hide technical complexity from its users, it often fails gruesomely. Windows users have come to expect that random failures, virus infections, lost data, and inscrutable configuration boxes are sad facts of life that one must put up with if one wants to accomplish things with a computer. For many of these people, life on the Mac side where things are promised to be easier is quite appealing.

The nature of Mac users and their expectations explains the difference in Apple’s product line and that of all other major manufacturers. Apple’s cheapest Laptop costs $1100, and it comes with the same version of OS X as all other Macs, as well as the same premium software. By contrast, Dell’s cheapest laptop is under $600, comes with Windows Vista Home Basic, and is preloaded with all kinds of nagware and spyware used to offset the ridiculously low price. If one wants “luxury” features like a version of Vista that supports the major features (Home Premium), enough RAM to support that version of Vista, a hard drive that doesn’t start out half full, Microsoft Word, and Antivirus, the cost has ballooned to around $900. Without these “luxury” features, the PC is almost useless; with them, the cost is reasonably close to that of a Mac that already comes with most everything necessary, and the PC still doesn’t come with a built-in camera or a FireWire port, and it’s a pound heavier and half an inch thicker.

The major manufacturers prey on the sensibilities of Windows users who do not believe that they need quality, convinced that because they perceive their needs to be modest, their computer should be the same. These are people who buy sub-$1000 computers, only to throw them out two to three years later when, through no fault of their own, their systems are overrun by spyware, adware and viruses and run as slow as molasses. They are simply unaware of superior alternatives or the advantages of using them, or are hesitant to try them out of fear of leaving their Windows comfort zone.

It is a shame that rather than nurturing and supporting these non-technical people who have no desire to become technical, Microsoft and PC manufacturers have set the standard to computers that often break down, fail to work, and in spite of everything, require significant technical knowledge to keep up and running smoothly. The appeal of Macs is that despite the fact that some technical knowledge is preferred, the baseline of quality will support the most non-technical users for a long time.

^[1]
Skype pictures brutally ripped off the Skype website and doctored for personal gain without permission. Sorry, Skype team.

^[2]
Note that this does not include those who have been coerced by their spouses/children/businesses/universities; only those who willingly made an informed decision can reasonably be counted.

^[3]
So says Time Magazine

More about the Dock

I got some strong reactions to my proposed version of the Dock the other day––so I’m going to address the most common ones here.

There’s nothing wrong with the Dock!
Yes, there is. I’ll admit that the Dock as it exists today is powerful, but its limitations are frustrating. Many of these limitations are fairly minor––mere annoyances at best––but it’s these minor aggravations that hold back a powerful system. I would say that Windows XP suffers from just this problem: it’s extremely powerful and versatile, but its lack of polish makes it difficult to unlock its power without being annoyed by stupid problems.

Besides, when it comes to interface issues, I always see the glass half empty. It’s not enough for me that the Dock does cool stuff. It has to do cool stuff and provide an obvious, consistent interface to users. For example, where someone else might say “I love how you can put folders as well as applications on it,” I would say, “It frustrates me how folders have to be on one side and can’t intermingle with applications. I’m perfectly capable of organizing my Dock items myself, and new users who try to drag a folder to the applications section and don’t see the icons shifting out of the way to make room for it might never think you can do it at all.”

It’s ugly!
How so? Does it look ugly because the buttons are square? Is the white too white? Not enough transparency? While pinstriping, shadowing, rounded widgets and transparency have been Apple’s hallmark for the past few years, they’re gradually phasing them out in favor or more mature interface elements. The first few versions of Mac OS X included a lot of transparency and shadowing for no real reason other than because it looked cool. In subsequent versions, transparency has been more and more sparse, and rounded interface elements are also undergoing the same transition. Take a look––here’s a conventional rounded pop-up menu compared to one from Mail.app:

Conventional rounded pop-up menu

Mail.app pop-up menu

Another example: here are the buttons from the Finder’s “Connect To Server” window:

Conventional rounded buttons
There’s really no point is providing an example, as rounded buttons like these are everywhere. However, a new guy is giving Mr. rounded button a run for his money:

Craaaaazy square buttons!

Now, whether or not you like square interface elements is really a matter of taste. There’s nothing inherently better about square buttons compared to rounded ones, but I personally think they look more mature and developed. As usual, Apple doesn’t seem to know what it thinks about the subject, which is why we have both rounded and square interface widgets. It seems to me that the logical progression is the squaring of rounded interface widgets, if only for the sake of change that Apple is so fond of. But whatever, people are entitled to their opinions.

It insults the user’s intelligence
Really? This was exactly the argument used by proponents of DOS and UNIX who scorned the concepts of mice and windows. Yeah, with enough time, users can get used to anything, but why should they have to? Better to have interface concepts that are apparent and consistent. I don’t see what’s insulting about making something that behaves like a button look like one. As for the arrow buttons, we’ll get to that later.

It doesn’t look good on the bottom of the screen
This is about those damn arrow buttons, I know it.

The arrow buttons don’t follow Fitt’s law
Aargh! Indeed they don’t!

Contextual menu items are so seldom used that the arrow buttons don’t even deserve to exist
Y’know, maybe they don’t. They do take up a lot of space, and really, right-clicking on a big square button is a lot easier than left-clicking on a small vertical one. Sigh. As much as I hate to admit it, it would be better for everyone if the square buttons were right-clickable. And really, there are other interface elements that are both right- and left-clickable, with each click doing different things. Some of them are even mildly buttonlike, such as Safari’s bookmark bar buttons. So to death with the menu buttons!

The eject button is too easy to misclick
Yeah, you’re right, and the Trash would be too easy to empty, too. As long as I’m getting rid of the arrow buttons, I might as well get rid of the Eject button.

Where that puts the dock
Here’s what the resulting Dock would look like:

Any better?

What should be done about the Dock?

The Dock bugs me.

It’s a launcher, it’s an application switcher, and it allows limited manipulation of apps, folders, windows, and anything else you put on it. It does an awful lot, and right now its interface is straining to accommodate all these tasks. A problem arises because some of these tasks require interface conventions that are diametrically opposed to others, which causes untold amounts of confusion when they’re smushed together without regard for the consequences.

The specific problems were outlined in the last post, but essentially the biggest one is that Dock icons only need to be single-clicked to be opened, even though they’re identical in appearance to their representation in the Finder, where they need to be double-clicked to be opened. But they can’t appear as buttons, because they need to be right-clickable, and nowhere else do buttons do anything when right-clicked. This fundamental conundrum is what drives the Dock’s usability problems.

So anyway, what to do about this? Well, in my mind, any icon that will do something when single-clicked should always look like a button. Maybe a little like this:

Two problems emerge. First, how will you be able to tell that Safari is open? At the moment, a little black arrow appears under the icon. This black arrow would look a bit silly in that button, so how about a stronger visual cue? I think that when Safari is open, the button’s background should darken, like so:

Well, that looks okay, and it brings the button more in line with other buttons that look similar when they’re pressed. But now we have a bigger problem: this button thing does not look right-clickable in the least bit. Well, let’s add an obvious visual cue that something can be done with this icon:

This little button with an arrow inside is obvious and easy to click on. The reason it’s to the right of the Safari button is because I like my Dock on the left-hand side of the screen rather than the bottom, but presumably, if this hypothetical Dock were to be placed on the bottom, the arrow button would be above Safari’s button and point upwards.

When you click on the arrow button, the menu you would expect pops up:

This solves the clickability problem: Dock icons are now buttonlike. And it makes the contextual menu functionality more obvious and easier to access. Mission accomplished, right?

But why not go farther? The Dock has always struck me as an interface element with so much power and potential; let’s see if we can harness it.

First a detour. I, and many other Mac users, have always liked that removable devices such as CDs and flash drives show up on the Desktop when you insert them or plug them in. In Windows, you have to go to My Computer to find them, and even then, they’re named with a seemingly arbitrarily-determined drive letter, and its name will often not show up or appear in all caps.

There’s no question in my mind that the Mac approach is more user-friendly, but it too can be confusing. If a flash drive shows up on the desktop, how come if you manually navigate to your Desktop folder (your username/Desktop), it doesn’t appear there? The ugly truth is that removable media (as well as internal media, for that matter) actually appear in some cryptic UNIXy location deep within the bowels of your filesystem where most sane people would rather be lit on fire by evil jesters then poke around in.

The reason that these volumes show up on the Desktop is because Mac users are used to this behavior; in pre-OS X days, the Desktop was where volumes showed up, and Apple kept the convention for familiarity’s sake.

Unfortunately, in maintaining this convention, OS X’s representation of volumes is confusing. Are they part of the filesystem? Do they live on the Desktop? If not, why do they appear there? If so, why don’t they show up in the Desktop folder? Do they behave like folders? Can they be manipulated like them? These questions all have ambiguous answers.

There should be a happy medium between Windows’ more technically correct but frustrating solution and OS X’s more familiar and obvious but technically problematic one. How about this as a compromise: volumes will show up in the bottom portion of the Dock. This way, they don’t confusingly clutter up the Desktop, but they’re also easily accessible at any time. Now, what are the two most common things users do with removable volumes? Open them up and eject them. Opening a volume is accomplished by clicking on it, and while an “eject” command could be placed in the contextual menu, how about just putting an eject button there to begin with?

I think this approach is superior to volumes appearing on the Desktop because the Dock is always accessible, and thus volumes that appear on it are also always accessible. And for those who complain that the buttons are too big, they can be resized just like before if you want to cram 20 or 30 apps into the Dock.

Of course, we can’t forget about the Trash. Its integration into the Dock has aggravated many prominent Mac geeks, and I worry that putting volumes on the Dock as well would cause undue strain in the Mac community, but if you think about it, it really makes quite a bit of sense to put it there. In my ideal model, everything else that traditionally goes on the Desktop is relocated to the Dock. One of the biggest problems at the moment is that out of all the traditionally Desktop-bound icons, only one wound up elsewhere, seemingly for no reason: the Trash. It would be a great boon to consistency if all of them went on the Dock. As with before, since emptying the Trash is the most common thing you generally do with it, why not make its button do just that? There wasn’t really any “empty Trash” symbol, so I dug around in the Character Palette:

While we’re at it, why doesn’t the Network sit on the Dock as well? It would certainly make it more accessible, as well as reinforce the concept of it being outside the local filesystem. Finally, the boot volume could go at the very bottom, separated from removable volumes. Remember, the actual location where these volumes show up is elsewhere in the filesystem where most people aren’t interested in exploring. Rather than additionally showing up in the Desktop, they would show up in the Dock instead. You could of course go to a volume’s actual location on the filesystem by clicking on the menu button and selecting the “show in Finder” command. But the point is that you don’t have to, because the Dock holds a handy representation of it, just as the Desktop does today.

The whole Dock, complete with applications, folders, removable volumes, Trash, network, and boot volume would end up looking something like this:

As a bonus, the top and bottom edges of the Dock butt up against the menubar and the bottom of the screen, respectively. This makes use of one of those awesomely-useful screen corners by putting the boot volume’s Dock button directly under (in this example) the bottom-left-most pixel, making it a breeze to click on it. Better yet, in the Dock’s preferences, you should be able to customize where apps, folders, and volumes appear on the Dock, so if you wanted, you could arrange it so that your Home folder or your most frequently-used app could be at the bottom left corner.

So there you go, a shiny new Dock. Whaddaya think?

Clickability part 2

Here’s another clickability issue: menus. Menus are completely universal in the Mac world; every application has them and uses them, and Mac users tend to be more comfortable hunting and digging around in menus than Windows users do simply because they’re unavoidable on the Mac. Unfortunately, there’s a frustrating clickability problem with them. Some menus have separator bars between groups of similar menu items, like so:

There’s nothing inherently wrong with these light gray separators. In fact, they help distinguish groups of related menu items from one another, which is generally a good thing. But the problem is what happens when one of these bars is clicked on.

In the menu above, the cursor is currently hovering above the “as List” item–you can tell because the whole row highlights blue.

When the mouse moves over a separator bar, however, nothing turns blue. This is as is should be, because nothing will happen when a blank area is clicked. Or so you think! In reality, clicking on a separator bar or a blank area in fact collapses the whole menu! This makes it easy to accidentally get rid of a menu if you vertically misclick. Yes, there’s a visual reminder that your cursor isn’t on a menu item, but why punish users who miss? It doesn’t make any sense to make them go back and re-open the menu.

Disclosure triangles
And now another clickability issue to illustrate the point: disclosure triangles.

They’re small, hard to hit and have unclear borders. Finally, because they’re gray–a color used for interface widgets that can’t currently be clicked– they look constantly unavailable.

Perversely enough, OS X already has a solution to this problem: put it in a button!


Say it with me now: “buttons are good, buttons are good…”

The Dock
The Dock is another perfect whipping boy for the clickability problem: it holds icons identical to those found in the Finder, so, in a perfect world, they should behave identically, right? Here Mr. Dock proves just how much of a jokester he is, because this concept is made a laughingstock by the Dock’s inconsistencies.

How do you open a Finder icon? Double click on it! So, when confronted with something visually identical to a Finder icon, what do you think most users will try to do to open it? Yup, you guessed it, they’ll double-click on it.

This of course accomplishes nothing for the Dock, as Dock icons only need to be single-clicked to open. The reason is that the Dock is a launcher–icons are put on the Dock to gain easy access to them. Probably 95% of the clicks the Dock receives are to launch icons on it; it seems unnecessary to require two clicks rather than one to accomplish this.

This differs from Finder icons which must be double-clicked to be opened; single-clicking only selects them. Why? Well, the Finder’s primary purpose is allowing icons to be manipulated–moved, copied, renamed, etc. These actions all require that their target be selected first.

But the Dock also allows for limited icon manipulation. In addition to launching things, the Dock can used to control an application in a limited fashion by right-clicking (control-clicking, secondary-clicking–whatever the kids are calling it these days) on its Docked icon. So, like Finder icons, Docked icons can have specific actions applied to them by right-clicking to pop up a contextual menu.

Here’s a summary of the current mess:

Unfortunately, making the Dock clickably consistent as it exists today opens up a whole can of interface worms. Essentially, the Dock does so much that it has more uses than there are interface conventions to cover them. The Dock is trying to be an application launcher, an application switcher, a folder opener, and at the same time provide a method for controlling applications.

As a launcher, ideally things that get opened when single-clicked should appear as buttons to denote their single-clickability. However, if Dock icons are also to be right-clickable (to control applications), it doesn’t make sense for them to appear as buttons, since buttons are never right-clickable.

So if you can right-click on it, it has to look like an icon. But icons need to be double-clicked to be opened, which is inefficient if the Dock is to be used to launch things quickly.

At the moment, there just isn’t an interface widget that appears as a button but also accepts right-clicks. Apple’s solution of just using the standard icon is overly confusing and inconsistent.

However, there are buttons that pop up menus when they’re clicked-and-held; these buttons are indicated by a little black arrow on the bottom-right. It’s possible that Dock icons could use such icons, but it doesn’t address the problem of right-clicking on a button. Oh well, maybe Apple could break the convention just that once. Or maybe, just maybe, Dock icons could require double-clicks to open.

Clickability part 1

Clickability refers to how obviously clickable interface elements appear to be, how many times they look like they should be clicked, and how well they respond to being clicked.

Many interface widgets in OS X unfortunately exhibit severe clickability problems. Much of the issue stems from the fact that the widgets rarely follow established guidelines, and sometimes those guidelines even encourage irresponsible interface design. Take this toolbar, for example:

Where to click? Are those icons even clickable at all? How many clicks do they require to activate? Can you drag the window by the grey area around the icons, or do those icons’ borders extend horizontally as far as the text beneath them?

The answers to these questions are not apparent without experimentation. For example, if you click on a patch of grey area and find that you’ve triggered a toolbar button, you learn that the grey area is part of that button. But consistent interfaces were designed to reduce the necessity of this experimentation to begin with. Busy people and non-techies don’t have time for experimentation and tend to become frustrated if they have to do it. I know–my parents are both intelligent, accomplished academics who are incredibly impatient with interface issues (though they–like most people–would never be able to phrase it as such). If they click on something once and it doesn’t produce the desired result, they’ll call for the nearest techie they know–usually me. if toolbar buttons had well-defined borders and looked clickable to begin with, this problem wouldn’t exist. Look how much clearer this would be:

Apple’s Human Interface Guidelines does not encourage the use of toolbar buttons that actually appear to be buttons–such as the ones illustrated immediately above. Instead, it promotes the use of icons, seemingly unaware of the fact that bare icons appear less clickable, less button-like, and have unclear borders. The Human Interface Guidelines is an irresponsible document! Bad HIG! Sit in the corner!

Another core tenet of clickability is that clickable things respond visually to being clicked. For example, in the finder, icons get grey boxes drawn around them when clicked. Toolbar buttons and Dock icons temporarily darken, as do most single-clickable buttons. And so on.

But there are some rogue interface widgets that absolutely refuse to follow the rules. I’m talking about titlebars, resize handles, and scrollbars. Moving and resizing windows and scrolling are three incredibly basic and common actions in any operating system. So why don’t the widgets that control them change their appearance somehow to show that they’ve been successfully clicked? Sure, once you move the mouse, you’ll know you aimed right if the window changes. But there’s no way for you to know if you missed unless you over- or undershoot the clickable area and end up mistakenly clicking something else–and then having to go back and fix it. Aaargh!

There seems to me to be no reason why titlebars (and toolbars, if unified), scrollbars, and resize handles don’t acknowledge a click by visually changing somehow. Since the standard seems to be a darkening of the clicked element, why not do this for those three? It would be perfectly natural-looking for scrollbars, which in fact had that behavior in the Classic Mac OS and continue to do so in Windows. It would also look fine for resize handles, and while a whole darkening title/toolbar might be a bit jarring at first, it would certainly provide excellent visual feedback!

How about another clickability problem: pulsing blue pushbuttons.


Quick quiz: is the user holding down the mouse button on the “Don’t Save” button? How about this next picture–can you tell which button is being pushed?

Answer: in the first picture, “Don’t Save” is being clicked. In the second picture, nothing is being clicked. Obvious, huh?

Those pushbuttons exhibit a number of usability and clickability problems. First of all, when a button is clicked on, it turns blue. Why is this a problem? Because the default button (the one that will be pushed if the user hits the return key) is also blue! In the first picture you can see the problem; when the user clicks on “Don’t Save,” the “save” button ceases to be blue! Why? Because the appearance of default buttons and currently pushed buttons are identical.

There’s a further problem with this behavior. Say you click on the default button–”Save,” in these examples. It starts out blue… but selected buttons also turn blue, so how can you tell if you’ve successfully pushed the save button? Well, it’s often ambiguous. Default buttons pulse until they’re clicked, so if you successfully click on one, it’ll cease to pulse. However, one particular frame of the buttons’ pulsing animation is identical to the “selected” shade of blue! This means that if you click the button while it’s in the middle of a pulse, it won’t change its appearance at all; it’ll only stop pulsing.

Most mortals simply don’t have time for such nonsense. When your average not-too-tech-savvy user pushes a button, she wants to see that it’s been pushed. This bizarre “pulsing-blue” changing to “not-pulsing-blue” behavior causes her to hesitate. Interfaces were designed to reduce hesitation through consistency. This button mess is counter-productive and entirely unnecessary. Sometimes it’s okay to sacrifice a little bit of utility for prettiness. This is not one of those times.

In pursuit of an excellent interface

“The secret of success is consistency of purpose.”
-Benjamin Disraeli

Why are good interfaces important? They’re the lens through which we view our computers, and they’re the control panel we use to interact with them. Because interfaces make up the layer directly between us and our machines, they are crucially important to our ability to use those machines efficiently.
A powerful computer capable of the most impressive feats of computation–but with an interface so Byzantine that its capabilities remain buried–is effectively useless; a decade-old clunker whose interface was designed with usability in mind can remain in service for years.

In fact, I am typing this introduction on an 8 year-old PowerBook G3. It has a passive-matrix screen, its battery holds approximately 8 seconds of charge, and it has no USB ports. By modern standards, it is laughably underpowered, and the eBay community seems to agree; I picked this machine up for $82. But despite its age and feebleness, it runs Mac OS 8, which was well-designed when it was released in 1999, and my black behemoth still feels relatively modern. In fact, as I navigate my hard drive, I occasionally come across a feature or two that I wish Mac OS X had.

Sure, OS 8 has its share of annoyances; every operating system (OS) does (naysayers will point to cooperative multitasking, lack of protected memory, and other buzzwords). But at its core, OS 8 was designed with the vision of usability and simplicity, using familiar metaphors, such as (gasp!) buttons.

In OS 8, things that open when you single-click on them appear as buttons, and things that open when you double-click on them highlight and darken on the first click. The clickable area of an icon is determined by the bounds of its picture. Clicked window widgets acknowledge clicks by changing their appearance. Dialog boxes use active verbs in buttons and can only be dismissed by clicking one of them. These conventions may seem simple and obvious, but most “modern” operating systems break many if not nearly all of them.

For example, in Windows XP, desktop icons can be selected by clicking in arbitrary areas around them because their clickable areas are seemingly random. In Mac OS X, many toolbar buttons don’t look anything like buttons and have unclear borders. In Linux, dialog boxes often have multiple ways to answer in the negative, with ambiguous results for each different method. Mac OS X’s so-called tabs don’t look anything like tabs. Windows makes you click a button marked “Turn off Computer” to restart it or make it hibernate. OS X scrollbars never look like they’ve been clicked. Many windows in Windows (haw haw haw) don’t show up in the Taskbar. And so on. The state of “modern” interface design is actually somewhere in the digital stone age.

Though this series of posts focuses specifically on the interface sins of Mac OS X, this isn’t to say that other operating systems fare better. In fact, OS X is by far the least bad of the other two most popular ones—Windows and whatever flavor of Linux you prefer. Volumes many times the size of this one could be written about the egregious interface transgressions of Windows. I focus on OS X for two reasons: both because it is the OS I am most familiar with, but also because it comes so close to excellence that I really don’t think that it would take entirely too much work to perfect it.

To business

For all the beauty of its downright lickable interface, Mac OS X often sacrifices uniformity in pursuit of this beauty. Here’s an example. The following are all toolbars on various applications that come with OS X:

Here’s what all toolbars used to look like in Mac OS X. Clear title bar area, pinstripey toolbar shelf, and toolbar buttons are big labeled icons. Simple enough that ColorSync Utility got to use it. Woohoo.

Here’s a new style–commonly known ad “Unified”–in which the titlebar area grew downward and engulfed the toolbar shelf, as seen in Xcode here People seem to like it.

System preferences uses this style (a variant of Unified), which consists of buttonlike buttons without icons that cannot be dragged around.

This toolbar belongs to Safari, which uses a bastardized form of the standard toolbar; when you hit the “customize toolbar” button, the resulting sheet shows buttons that are clearly labeled, yet when you drag them to Safari’s toolbar area, the labels disappear!

Even though these buttons of iTunes’s look like they might belong to a toolbar, they don’t: those buttons are immobile as your dead granny.

Mail uses these odd toolbar buttons, which look different from standard buttons for no perceptible reason. On the plus side, they look more “buttonlike,” but on the minus side, they’re pretty fugly.

There is a very serious problem here. Though they all have the same purpose–to put frequently-used commands and tools in an easy-to-access location–none of these toolbars look the same. Why not? Some have obvious borders, indicated by the “lozenge” surrounding the words or symbols within them; some do not. Some lozenges are white; some are not. Some buttons have text indicating their function beneath them; some do not. Some windows have a clear line separating the titlebar and the toolbar; some do not. Some toolbars are darker than others; some are not.

And so on. These toolbars and buttons are all pretty, but none of them is consistent with any other one, and none of them appear to follow concrete guidelines that might govern what they should look like. Apple’s own Human Interface Guidelines–the supposed de facto standard for OS X’s interface–is rarely taken seriously anymore, it seems–even at Apple! It’s this sort of seemingly random inconsistency that those guidelines were written to prevent, and for good reason: an inconsistent interface is confusing to new users.

OS X is plagued by this irritating inconsistency, which manifests itself through a couple of big, broad issues that OS X’s interface designers seem to have forgotten about, and which I will address in future posts: clickability, predictability, and adherence to real-life metaphors.

« Previous Page

• Recent Posts

  • Microsoft’s Murderous Partnerships
  • Following the money, losing the way
  • Snowballs and dominos
  • Fun in PC land, or why Macs are faster
  • Must slam head into wall

• Blogroll

  • Chameleon Insurrection
  • Chase of the white hart
  • Crazy Apple Rumors
  • Daring Fireball
  • RoughlyDrafted

Blog at WordPress.com. | Theme Light by Stephen Reinhardt