Tech Talk with David Pogue

Wee Mousie, Fear Not a Touch PC

Blame “Minority Report.”

Or maybe the iPhone. Either way, multitouch computing is white-hot these days. There was just something so seductive about the way Tom Cruise could grab on-screen objects and toss them around — not with a mouse, but with his fingers.

What’s so great about the mouse, anyway? How is it natural to drag a plastic box across a horizontal surface to move things on a vertical screen two feet away? How kludgy! How indirect! How Cro-Magnon!

Toshiba Satellite M500

The iPhone brought multitouch to the real world. You want to push a button? So push it. You want to turn a page? Flick the screen like a real page. Move down a map by dragging with your finger, as you would a real map. Spread two fingers on the glass to zoom into a page or a photo, as if it’s printed on a sheet of rubber. Twist two fingers against the glass to rotate a photo. It’s direct, it’s logical, it’s addictive.

But why should multitouch be confined to your phone? We should have multitouch screens on our computers, right?

Microsoft has seen the writing on the glass. Right out of the box, Windows 7 recognizes all of those iPhonish finger gestures: pinch, spread, rotate, flick, tap, double-tap. You can even simulate the right-click of a mouse: hold your index finger on something on screen, then tap somewhere else on screen with another finger. All of this comes to life when you buy a computer with a multitouch screen. They’re just starting to appear.

Now, making gestures on a small phone screen, perfectly angled in your hand, is one thing; making gestures on a big, immovable, vertical PC screen is quite another. Keeping your arm out at right angles, making tiny, precise movements on the glass — wouldn’t that be an exhausting, ergonomic nightmare?

In fact, the aches and swelling that would supposedly result has earned itself a nickname online: Gorilla Arm. Fear of Gorilla Arm, we’re told, killed touch computing during its first wave in the early 1980s.

But is Gorilla Arm for real? After all, as the Wikipedia article on touch-screen computing defiantly observes, “fine art painters and draftsmen have worked in similar postures with vertically mounted surfaces to draw on for millennia.”

So what’s the truth? Is multitouch computing the wave of the future? Or just an expensive gimmick that will send us all to the physical therapist?

I decided to try it out for myself. Toshiba lent me a touch-screen laptop, the Satellite M505 ($1,000); Hewlett-Packard sent me a touch-screen all-in-one desktop computer, the TouchSmart 600 ($1,170).

The TouchSmart is an all-in-one PC (the computer itself is inside the screen). Snazzy features are everywhere: wireless keyboard, wireless mouse, tilting webcam, memory card slot, adjustable stand, even a slick light on the screen’s bottom edge to illuminate the keyboard. It has enough jacks to stock a RadioShack, U.S.B. ports by the handful, HDMI input so you can use the 23-inch hi-def screen for your home theater, and a jack for a TV cable so you can use Windows Media Center to act as a fancy TiVo.

Unfortunately, the TouchSmart may be loaded, loaded, loaded, but it’s also slow, slow, slow. Slow to start up, to open programs and, tragically, to respond to finger taps. H.P. confirms that it’s made with “mobile components,” meaning laptop parts; in any case, it feels overwhelmed by the simplest tasks.

When you spread two fingers to zoom a Web page, the zooming animation is jerky and therefore hard to control. You can swipe horizontally to go back or forward a page, but you’ll wait — no joke — nine seconds for the computer to respond.

The Toshiba laptop is another story. It’s a rather thick, heavy laptop with flat, slippery keys and trackpad buttons — but the touch screen is fast and responsive.

Here, you get giddy with the possibilities. A button on the screen? Just tap it. Reaching for the mouse and guiding the cursor all the way across the screen feels so much less natural. A menu or list? Just drive directly for the item you want, rather than passing through the whole list with the arrow keys or the mouse.

Unfortunately, though, there are four obstacles to making multitouch work as well on a PC as it does on an app phone.

First, it has to work every time. The H.P.’s touch response is so balky that you tap and wait to see if the PC “heard” you. (It doesn’t have the touch-sensitive surface of app phones and multitouch laptops because that would cost too much on such huge screen. Instead, it has tiny infrared cameras mounted at the edges of the screen. High-tech, but not sure-fire.)

Second, there’s the finger-size problem. On an app phone, every tappable element on the screen is big and finger-friendly — not so on a PC.


Now, Windows 7 helps out in all kinds of small ways. On a touch-screen PC, the Show Desktop button is double-wide; you can open jump lists (pop-up taskbar menus) by swiping a finger upward instead of right-clicking. If you’re clever, you can customize close boxes or scroll bars to make them larger.

In Internet Explorer 8, you can tap with two fingers to zoom in, which is handy when you want to magnify the area of a link. The History and Favorites lists are more widely spaced.

But the rest of Windowsland remains as it was: mostly too small to tap with your fingers.

Hewlett-Packard TouchSmart 600 PC

H.P. and Toshiba each compensate with special touch programs, like drag-your-photos modules and scribble-a-note apps. H.P.’s is especially ambitious. It’s a mini-world of mini-apps: Photo, Weather, Clock, Live TV, DVD, Calendar, R.S.S. Feeds, Browser, Video, Recipe Box, Twitter, Netflix, Canvas, Webcam, Hulu (great Internet TV) and so on.

But H.P. and Toshiba can’t rewrite every app on earth. Until somebody does, you’re going to experience a mishmash of touch-appropriate and too-small software.

The third huge problem: today’s programs don’t respond consistently to touch gestures.

In Microsoft’s own Windows Live Photo Gallery, the right-click gesture works in thumbnail view, but not once you’ve opened a photo. In Adobe Reader, scrolling works and spreading works, but not two-finger-tap-to-magnify. The rotate gesture works in Photo Gallery, but not in Google’s Picasa. (Even the swipe-to-scroll gesture doesn’t work in Picasa.)

Touch is nearly useless when you’re word processing. You can zoom in, and you can double-tap to select a word, but precision dragging (for highlighting a passage) is nearly impossible.

The final problem: our old friend Gorilla Arm. After an afternoon of using arm’s-length touch gestures, my wrist was tingly and sore. Millennia of draftsmen may have worked in that position, but they probably needed a lot of Advil.

Now, when everything’s working — when you’re in an app that reliably responds to the gestures — it really works like magic. You flick, zoom, manipulate and wonder why anyone puts up with the mouse.

The trouble is, those moments are far too infrequent. You never know when a gesture is going to work, so there’s a lot of trial, error and feeling foolish.

Now, the iPad and its inevitable imitators are a different story. If you can hold the thing at any angle, if the hardware and software were designed for finger operation from the beginning, multitouch can be far more successful.

Multitouch regular PCs, however, are a long way away from being usable or pleasant. Until the consistency and design problems go away, multitouch will only complement, not replace, the mouse.

In other words, the depiction of a futuristic, mouseless, all-touch interface in “Minority Report” clearly skipped over a few important details. You never saw Tom Cruise making an appointment to see his doctor about Gorilla Arm.