Previous  |  Next


Dec 12

Tim O'Reilly

Tim O'Reilly

Multi-touch Raises the Bar

There's been an awful lot of commentary about Amazon's new Kindle. It's clearly lived up to its name, and ignited a lot of discussion and excitement about the new possibilities of the always-on electronic book.

I wanted to add one more piece of fuel to the fire. But it's not about ebooks, or even much about the Kindle. It's about the way that the iPhone has changed my expectations of electronic devices. When I picked up my first Kindle, I immediately began to stroke the screen. Nothing happened. The device felt curiously inert, almost dead. It took me a moment to remember that the controls were on the side.

I share this as a small bit of "news from the future." It's hard to imagine that by this time next year, portable devices without multi-touch screens are going to feel very old, and very quaint.

It's interesting to me that simple touch screens don't create this effect. But once you've worked with a multi-touch device like the iPhone, it's hard to go back.

Here's a question: how long after Apple brings books to the iPhone (or its successor) will it be before Amazon brings multi-touch to the Kindle?

tags: amazon, apple, iphone, kindle, nff, web_2.0  | comments: 17   | Sphere It


0 TrackBacks

TrackBack URL for this entry:

Robbie   [12.12.07 10:50 AM]

If you wanted a touch screen on the eReader you should have tought about the Iliad, or if you want something sexier (currently unfortunately without the touch screen) you should have thought about the Readius
Both are spinoffs of Philips and to me they look more promising than the kindle.

Iliad is on sale in Dutch book shops and it looks like the Readius will be on sale before Christmas.

They're not yet multitouch, but the Readius looks very promising to me, especially because it contains 3G capabilities.

William Donohue   [12.12.07 11:17 AM]

Two things made me believe multi-touch will become the interface for our more physical media (maps/books/etc.) - something Danny Hillis said at ETech 2005, and Jeff Han's multi-touch map manipulation demo. Danny mentioned attending a cartographer's convention, and the sense he got that the mapmakers missed the physical maps they had grown up with, despite the advantages of GIS and digital map systems. This slots into Jeff Han's map demo - it seemed to me that his team had gotten the gestural interface just right, and this made the experience of the map both more useful and more enticing. Digital maps could have better utility *and* be even more engaging than paper maps, coming down off the screen and back into our hands. I think something similar could happen for ebook readers, provided someone gets the interface right. Ultimately, the first tools we all use are our hands, and all our interface devices are just extensions. Multi-touch puts us into more direct, more immediate contact with digital media.

nick   [12.12.07 02:32 PM]

I've not used many touch or multi touch devices. We had a touch screen PC once and I found the lack of precision and the occlusion of a fat finger to be very distracting.

For something like an iPhone or a tablet does it make sense to put the touch on the back and the screen on the front? As you naturally cradle / hold the device you use gesture on the back face while the computing renders the touches appropriately on the screen. No occlusions, no finger smears on the screen. You can even overlay some wire frame / alpha modulated keyboard.

Don   [12.12.07 03:01 PM]

I hope the news from the future is right. I've used a Wacom pen tablet in preference to a mouse for at least 5 years now and my Nokia N800 Internet Tablet is screen touch based. There's been more than one groggy morning when I've approached my desktop PC with index finger oustretched towards the display. I crave a tablet PC (or Mac) but mobility PCing is provided by my employer so I've really no motivation to buy my own.

Paul Mison   [12.12.07 03:48 PM]

From the "out of the mouths of babes" department: when visiting the US this summer, I met up with a friend with an iPhone and an 18 month old child. I tried to show pictures using the screen on my digital camera, and the boy tried to scroll through them as if it was a touch screen. (It wasn't, and I expect digital SLRs will probably be the last cameras that don't have them.)

rketide   [12.12.07 05:40 PM]

what little i've seen on iphones multi-touch, it seems like a terrible waste of a good technology.

i do come from a background learn one handed keyboards and wear scifi flick data gloves on a regular basis though, so gesture interfaces lack some of the novelty value.

the reason i've always sweated multitouch is for collaborative work. the main case where you need more than one input pointer is having more than one user at the same time. collaborative technology demands multi touch.

we've started scratching the ubiquity future where data trumps device, but we're still locked in the oldschool of each person having their own device. i'd much rather share huge high res table sized displays with people and be able to shove documents across the table at them. heh, for someone who thinks the desktop metaphor has been holding back the computer technology, its odd to find a place where i wish the desktop metaphor better followed the real thing.

Roy Schestowitz   [12.12.07 05:44 PM]

The next version of X Server will have multi-touch support built in, so you could expected Kindle to add that.

rektide   [12.12.07 06:27 PM]

Roy: MPX (Multi pointer X) and Peter Hutterer are my heroes! I've been interested in multiple concurrent input systems since `98 and once downloaded the X source to try and start implementing it myself. I never got anywhere, but thank god, someone else with skill finally decided it might be important and lo and behold has a wonderful implementation.

Adam Hodgkin   [12.12.07 11:40 PM]

It will be interesting to see how much variety and consistency evolves in the (multi) touchscreen interface. Apple have made a great start (lets hope it doesnt get messed up the way the mobile phone numeric interface has been). Touching and gesturing to navigate is very satisfying. Surely this re-introduction of a somatic and gestural component to the way we read is a profound change.

Alain Pierrot   [12.13.07 01:30 AM]

Many things could be done, combining touching and gesturing. One mustn't forget the embedded accelerometer, which could yield new conventions: cf.

And some ideas could come from 3D interfaces such as BumpTop 3D:

Kyle Mulka   [12.13.07 02:54 AM]

I don't think Apple will bring books to the iPhone. They may make another device that's better for reading books though. I'm looking forward to a multi-touch tablet computer from Apple. I feel like its just around the corner.

I don't think multi-touch would go very well with the current generation of e-ink technology. You would drag your figure across the screen, then have to wait for a second while the screen refreshes. It wouldn't be a very good use experience. I think its better to have discrete events for refreshing the screen, such as a button press.

Tim O'Reilly   [12.13.07 07:57 AM]

Lots of good comments here while I was away yesterday. Let me try to respond to them all at one go.

Kyle -- if Apple is truly opening up the iPhone to outside developers, as they claim, they don't need to be the ones to bring books to the iPhone. But yes, I bet they'll do other devices as well. "Just one more thing..." And very good point about the incompatibility of e-ink and multi-touch.

Adam -- I agree. We'll probably see wacky overuse of multi-touch before interface conventions stabilize. But I think Apple got an awful lot right for a 1.0 device.

Roy -- fantastic news about multi-touch support in X. Obviously, for this next UI revolution to take off, we'll need widespread device support, but also a software stack. And having an open source software stack is going to be very important, since I imagine that there are going to be lots of closed (and patented) implementations.

rektide -- I was talking with Jeff Han yesterday, and he made the same point, that the real sweet spot for multi-touch is with big display/computing surfaces for collaborative work. That being said, I do think that the iPhone has set the bar for UI on portable devices. But Jeff pointed out that there will be limits that will keep this on high-end devices, notably power consumption. What makes the iPhone so attractive isn't just multi-touch, but also graphic performance, which means an onboard GPU, which takes power, etc.

Paul -- what you noted with the small child is exactly why I think multi-touch will become the standard. Once you've started to use it, it changes your expectations.

Nick and Robbie -- don't confuse touch screens with multi-touch. Multi-touch means that the screen doesn't get confused with multiple touches at the same time, and that there's support for gestures, like that fabulous flicking motion to scroll on the iPhone.

But again channelling Jeff Han, he sees a period of confusion, as many people with ordinary touch screens start claiming that they are multi-touch. What's more, some implementations are less powerful than others -- for example, with some edge-based sensors like those used in the great Danny Hillis demo of the Applied Minds map table that William mentioned, one touch can obscure another if they are lined up so that one blocks the view of the other. With an embedded grid sensor like the one in the iPhone, you get more resolution. But Jeff hinted (but wouldn't say more) that there are even more sophisticated systems (presumably like his own.)

So expect a lot of market confusion. And speaking of confusion, I heard that Apple is trying to trademark the term multi-touch. If they succeed, that seems like serious over-reaching, and a bad thing. I hope someone at the trademark office reads this blog, and realizes that multi-touch is a descriptive term that existed long before Apple introduced the iPhone!

Coleman Foley   [12.13.07 11:51 AM]

amazon would have to copy the iphone right away if it started having ebooks

rektide   [12.13.07 09:04 PM]

Tim, I'm very glad to hear I am not alone re cooperative being the killer app for multi touch. I will look forward to hearing more on the topic of cooperative computing environments from Han in the future. I've seen one of his demos with two people having all five fingers on the screen and was impressed, I hope he can carry on from that demo and do some real collaborative environments.

You've probably already seen it, but Johnny Chung Lee over at CMU built a multi touch display with a $35 wii mote.


MPX has an experimental window manager that has some really good techniques for assigning individual windows to individual (or groups of) people. X applications all assume a single pointer, sometimes you get really freaky behavior if theres three different cursors interacting with a window at the same time (usually in totally different contexts), so this can impose some extra rigor in collaborative spaces.

Re graphics, we're only just getting vector machines in consumer systems (so called stream processors from the newest nvidia and ati cards). ATI for example has 320 processors running 750 mhz. cpu power dissipation comes from two places: the fab process has inherent static losses from the resistivity of the semiconductor (hence hi-k dielectrics), and the dynamic loss, equal to the voltage squared times the switching speed. Another twelve years down the road this is going to hit the embedded space like a bomb: take a couple hundred cores, drop them to really low mhz, reduce the voltage to sub-volt levels and scale out massively. Each processor has to be really simple, preferably using shared control logic, which is the definition of a vector computer. Vector computers work in very few places for very specific tasks, but graphics are one of those areas where they work brilliantly.

I applaud apple for getting such great performance out of an ARM processor, and I love that they're sharing the love and contributing back into what has made it possible for them, LLVM.

rektide   [12.13.07 09:10 PM]

$35 wiimote multi touch:

a great start. :)

Greg Raiz   [12.14.07 02:03 PM]

It's not the multi-touch that makes the iPhone great. The thing that the iPhone has and the Kindle doesn't is great design.

Fluidity, Elasticity, Transitions, etc. These are designed in.

While components like multi-touch capacitive screens will become more popular most devices will continue to feel dead because most companies are not willing to spend years to make the designs come alive.

Michael R. Bernstein   [12.20.07 10:42 AM]

PyCon 2008 has a Python-powered multi-touch talk scheduled:

Post A Comment:

 (please be patient, comments may take awhile to post)

Remember Me?

Subscribe to this Site

Radar RSS feed