Thursday, December 20, 2012

Small is Beautiful: The iPad Mini

A month ago, I sold my 3rd-generation, Retina-screen iPad and bought an iPad Mini. It was inevitable, given my taste in equipment: I never owned an iPod before the Nano. My favorite pre-Intel Mac was the 12” PowerBook. Before 2009, I never even considered a MacBook Pro because they started at 15”. The day the 11” MacBook Air was announced I told my wife it was my next computer. And since acquiring the iPad—even smaller than the Air—it has been my main computer.

And therein lay my one reservation, the feeling that this time I may have gone too small. The Mini's appeal is obvious: greater portability, easier to hold in one hand while reading, lower cost of entry, sleek and attractive design. But my iPad functioned as a computer. I created and maintained spreadsheets on it. I used it to surf the web and write email. Prior to negotiating my current position, I even conducted an entire job search on it—finding listings via Safari, creating resumes and cover letters in Pages, uploading them to recruitment sites via iCab, and so on.

All of those tasks can still be done. In fact, thanks to the Mini’s small footprint, they can be done in even more situations. But I worried about losing another fundamental computing task: I can touch-type on the full-size iPad screen. Not as fast as with a physical keyboard, but fast enough to accomplish real work (with iOS’s sometimes-maddening but generally-excellent autocorrect, my iPad typing was also surprisingly accurate). I was certain the Mini would make touch-typing impossible.

If I put on my reviewer's hat, then I can see this fear is unwarranted. I suspect the vast majority of users who haven’t tried (and even most who have) feel that touch-typing is something you do with a keyboard, not a touchscreen. Losing that capability on the Mini is a non-issue. And, in fact, I have been pleasantly surprised by how well I can type on the Mini in landscape orientation. My speed and accuracy dropped, but my typing is still passable. Typists with larger hands will have a more difficult time, but the Mini offers a happy tradeoff—thumb-typing in portrait is easier than on the 10” iPads. And, naturally, the experience with a Bluetooth keyboard is just as good as on the full-size iPad.

“Just as good” evokes one thing all potential buyers should know: the experience of the Mini is the same as on the full-size iPad, only smaller. iOS is the same, the apps are the same, and the Mini even benchmarks about as well as the 3rd-generation Retina iPad. I find the reading experience to be better for ebooks and the web. This is a credit to its 7.9 by 5.3 by 0.28 inch size and its mere .68 lbs. weight. Apple’s vivid comparisons (“as thin as a pencil”) fail to do it justice. It truly feels too thin, too light, to be capable of what it can do. Which is, again, anything that a full-size iPad can do.

Except, of course, display Retina graphics. This has been a sticking point in commentary on the iPad Mini when announced (and reviews when launched), but the concern has been overblown. The screen offers the same resolution (and a higher pixel density) for those upgrading from a first-generation iPad or iPad 2. And surely the new form factor and lower price point will attract new buyers to the iPad. Those two groups will surely make up the majority of first-year buyers of the Mini. For those who do “upgrade” from the first Retina iPad to the Mini (as I did), I feel the concern is still exaggerated. I definitely noticed the difference. For about a day. The pixel density on the Mini is fine, particularly at the distance it is usually held from the eyes. After a few hours of use, graphics and text on the Mini look crisp and clear. I look forward to a Retina version, but the Mini is good enough I doubt I’ll upgrade immediately.

Speaking of overblown concerns, I think more can (and should) be said about the form factor. Size, after all, defines Mini. The standard iPad 3:4 aspect ratio means widescreen video will always be letterboxed, but the tradeoff is more than worth it as it makes for a good experience in most apps in either portrait or landscape (and video still looks great). Much has been made of how Steve Jobs slammed the 7” form factor of many Android tablet, and yet the post-Jobs Apple ships a smaller tablet. Of course, such flip-flopping accusations ignore the fact that the iPad Mini’s 7.9” screen is closer to eight than to seven, and that distinction makes a world of difference. To put it in terms an Android proponent will understand: if a 4”Android phone screen was an advantage over the 3.5” iPhone screen, then an 8” screen on an iPad Mini should be, proportionally, just as big an advantage over a 7”-screen Android tablet.

And the screen size is good for iOS; as Apple has said, if developers follow tap-target-size guidance, their apps will already be appropriately-sized for the Mini. I haven’t had much trouble with touch in general—the “accidental touch” detection in the edges works well (to prevent, for example, inadvertent page turns in iBooks when resting a thumb on the screen). I haven't noticed any added difficulty in hitting even fairly small tap targets (the red “delete app” X’s, for example). I don’t know if that would be true for a 7” or smaller iPad.

The size makes the Mini the best eReader I’ve owned, the best tablet I’ve owned, and maybe even my favorite computer I’ve owned. It is capable and fun to use—just like a full-size iPad—and almost insanely portable. I don’t know that I fully agree with Dan Frommer’s quip that “This is the real iPad,” but the Mini is definitely something special, and it's the only iPad for me. 

Thursday, December 13, 2012

Not a Mac, Not a PC: Is the iPad a computer?

Even the staunchest tablet computing skeptics will probably concede that yes, in the literal sense, the iPad is a computer. It has a processor, memory, storage, and runs software. But I suspect many would still cling to the notion that in some fundamental respects, the iPad is something different. Certainly we see this played out in market share statistics, which often count the iPad and other tablets separately from computer sales. But should they? Is the iPad a computer, or not?*

The language of tablets argues against the iPad-as-computer theory. The term “media tablet” often appears in commentary as a way of denigrating tablets as mere “consumption” devices. And the term sometimes fits. The Kindle Fire and Nexus 7 were pitched as gateways to purchase and enjoy content from Amazon and Google stores. Barnes and Noble called their first tablet, the Nook Color, “the reader's tablet” (and initially told developers to submit apps that “complement the reading experience.”). You can do creative and productive things with these devices, but for them the term “media tablet” is less a criticism than a mission statement. 

Does an iPad also fit this category? The term implies a distinction, suggesting the existence of some other type of tablet—a true “tablet computer,” perhaps. Which devices on the market today fill that imaginary category? The Microsoft Surface with Windows RT is a strong candidate. Surface ships with a desktop experience and traditional Office suite, after all. But I wager there is more productivity software in the iOS App Store than software of any kind in the Windows Store.

Perhaps the best argument in favor of the “media tablet” label is usage. In my estimation, iPads are used by most people, most of the time, for consuming content rather than creating it (or doing “real work”). But this is also the wrong argument. Usage is a clumsy way to decide what constitutes a “real” computer.

Think about non-techie family members, whose Windows desktops serve only for Facebook browsing and casual gaming. For that matter, ten years ago, my then-roommate—at the time far more tech-savvy than I—built his own PC tower, but used it solely for gaming and web surfing. Surely there are models of PCs, even entire brands, that appear at price points and in retail channels that appeal primarily to these user classes. Would you argue that such PCs are not real computers? Obviously not—even when mostly used for entertainment rather than work, they have the potential to do real work and create content.  But then, so does the iPad, which means it, too, is a computer.  

What behavior would you require of a computer before calling it a computer? Should it allow users to create, edit, update, and display Office-format documents, spreadsheets, and slide decks?  Surf the web, send email, and engage in social networks?  Create web content?  Watch videos and listen to music and podcasts?  Record, edit, and upload audio and video files? Print? Connect peripherals? Allow touch typing (or even allow for a physical keyboard)? Run software? Create software? Create software to run on the machine itself? If you define computer-ness by any of these tasks, then the iPad counts.

The last item on that list deserves further thought. A number of iOS apps now allow users to write and execute code. One example, Codea, was used to develop at least one game that is currently sold in the App Store. Granted, compiling and submitting such apps to Apple requires a Mac running XCode, but the software was developed and tested on the iPad. If the iPad wasn’t a computer because it couldn’t be used to write and compile its own software—as has been argued before—then neither was the original Macintosh a computer (which required a LISA). And with Codea, the iPad is closer on this score than the currently-shipping Surface tablet.

All arguments against iPad-as-a-computer that I have encountered fail for similar reasons to the “programmability test”: wherever the line is placed, it will exclude something that is clearly the computer, or include something absurd. Take multitasking, for example, a core feature of modern operating systems and user experience. Let’s set aside whether iOS’s background processing and efficient save-and-resume functions qualify as multitasking. Suppose we concede that the iPad doesn’t multitask the way a PC does. If multitasking is a mark of a true computer, wouldn’t that mean HP’s failed TouchPad is a computer? A major selling point for WebOS devices, including the TouchPad and Pre phones, was “true multitasking.” Do we consider the TouchPad—with its anemic software ecosystem and tiny user base—a computer, but exclude the iPad?

Why are the arguments for counting iPads separately from “computers” so leaky? I believe it is because the line is drawn based on a single criterion: it puts the iPad on the opposite side from Macs and PCs. If you define "true computers" in a way that deliberately excludes a device that is clearly a computer, your definition will by necessity let too much in or keep too much out.

All of which begs the question: why does anyone care about excluding the iPad from the ranks of “true” computers? Industry politics may play a role. Counting iPad sales among computer sales puts Apple way ahead, overturning a lot of conventional wisdom. But I think this is more than factionalism. The notional “media tablet” category makes it easier to think about what it means to have powerful, usable computers that are not Windows PCs (the comfort-zone answer: “Nothing. Real computers run Windows.”). If the iPad, and tablets generally, are computers, then computing itself is changing in fundamental ways. That's a bigger story than iPad-as-consumer gadget. It’s harder to think about, write about, or make decisions based upon this idea. Which is probably why so few make the leap—and why this essay is on whether the iPad is a computer, not what it means if it is.

* - Note that I am not, as many past articles have, asking if the iPad is a PC. To whatever extent the iPad is a computer, it is surely a very “personal” computer, but “PC” has largely become synonymous with “Windows” (for better or for worse). (back)

Thursday, December 06, 2012

No, Compromises: The Microsoft Surface

 “Oh, my God, we have so, so many of them.”

The exasperation in the associate’s voice told me all I needed to know about Surface sales—this was not a supply chain success story. No, the Microsoft store had a lot of Surface tablets, and it wasn’t supposed to. Digitimes reports Microsoft cut component orders by half, Piper Jaffray recorded zero sales in a two-hour observation period on Black Friday, and the blustering Steve Ballmer even called sales “modest.” After my own two hours playing with and testing the tablets, I haven't seen a single Surface sold yet, either.

This disappoints me because the Surface with Windows RT (and Windows 8 generally) fascinate me. They offer an interesting and unique vision of touch computing (if not of “tablet” computing). I believe that good design is almost always opinionated—even the “right” design, or a design with broad appeal, will say No! to some ideas and some users. Apple clearly knows this; they prioritize design, and remain polarizing even as they grow ever more popular. Microsoft on some level gets this, as the new Windows overflows with opinions. But it also fails to deliver on Microsoft's “no compromises” boasts. 

The Surface feels boxier than the iPad, echoing the distinction between iOS’s rounded edges and textures and Metro style’s flat, spare square designs. The thickness, squared edges, and elongated design seem to give it more physical presence than its 1.5 lbs suggests (it doesn’t help that I first held a Surface and an iPad mini on the same night). The magnesium alloy shell has a different character from the iPad’s carved aluminum, but feels pleasantly solid. The screen looks better to my eyes than pre-Retina iPads, and closing the kickstand or attaching the Touch Cover result in satisfying clicks and snaps.

The kickstand and the central role played by the Touch and Type Covers highlight the opinionated design of the Surface. In contrast to the various laptop docks pioneered by Asus and others, these keyboards don’t physically support the screen, and the kickstand works best on a flat surface (pun intended). The Surface almost becomes a highly portable desktop computer, rather than a laptop replacement or a tablet. Other factors constrain tablet-style usage, as well. The placement of the Windows button under the screen suggests landscape orientation; the angles and placement of the kickstand, cameras, and keyboard connection demand it.

Video looks immense on the 16:9 aspect ratio screen when compared to a letterboxed view on the 4:3 iPad. But holding the Surface aloft in portrait mode feels awkward—the content becomes too tall and narrow for comfortable reading, and the size and shape are even less one-hand-able than full-size iPads. That said, the widescreen aspect ratio looks more clever here than on 16:9 Android tablets. In the Surface’s main usage scenario (with keyboard and kickstand), a 16:9 shape minimizes the distance between the keyboard and the top of the screen. This limits arm fatigue from reaching out to tap a vertically-oriented screen.

And users will definitely want to reach out to the screen. Windows RT includes a rich gesture language: swipe down through the bottom of the screen to close an app, in from the right to trigger the search and sharing “charms,” or from the left to switch apps. I worry about discoverability—I only found the gestures to “snap” a second app alongside the first through trial and error, and only tried because I knew the feature existed. Scrolling and zooming were smooth via screen but jerky via the Touch Cover trackpad. Keyboard shortcuts rarely worked on the first try.

Clunky or not, the keyboard covers help with the included Windows desktop and Office suite. Here, the heavily-opinionated Start Screen experience and physical design give way to the classic Microsoft approach: include everything, well-done or not. Office apps all launched quickly and felt responsive enough, and did everything you expect (unless you expect to run macros). However, window controls and Ribbon icons haven’t been visibly redesigned to be touch-friendly, and I had a little trouble placing the cursor or selecting ranges of cells. The apps are useable, but leave me wondering why there isn’t a true touch-first Office UI.

Other apps were fine, despite some glaring omissions. I didn’t road-test directions, but the maps app worked smoothly—in a few seconds of pinches and swipes, I found my house in Kansas and Disney World in Florida. The weather app was attractively flat and Metro-styled, a strong contrast to its iOS counterpart. Switching among apps was fast and easy, and the ability to view two apps side-by-side seems like a strong competitive advantage (making good use of the 16:9 screen). Angry Birds Space and other apps downloaded from the Windows store all worked as well as on other tablets.

But is that enough? I use Windows all day at work and an iPad as my primary computer at home, so the Surface sounds great in theory. But I can’t justify buying a Windows PC of any stripe that can’t run Microsoft’s Visual Studio. I am also going on four years as an iPhone power user, and iTunes would still be handy (and would be a must for my wife, if we bought a Surface for her). As of today, the Surface RT can’t run these—or any—legacy Windows apps.

True, iTunes barely runs on today’s Windows hardware and creating a useable ARM-based version  of Visual Studio 2012 probably constitutes a massive engineering feat (if it’s even achievable). The x86 “Pro” version is coming soon, but I don’t want to pay the Intel penalty in thickness, weight, heat, noise, battery lifeor price. Do I ask too much? Perhaps, but it was Microsoft’s Steven Sinofsky and Steve Ballmer who said of Windows 8, “No compromises.” Yet here we are, unable to perform core computing activities on the current Surface—and coping with a large variety of weird design tradeoffs. Surface shifts the emphasis—not “No compromises,” but rather: “No, compromises!”*

* - Credit where credit is due: I stole this joke from The Simpsons. When confronted with an ad that reads "WORKS ON CONTINGENCY   NO MONEY DOWN," shady attorney Lionel Hutz claims it's a typo and adds punctuation to change it to "WORKS ON CONTINGENCY? NO, MONEY DOWN!"