Thanks to everyone who attended.

JZ

>>>>>>>>>>>>>>>>>>
Rocky Mountain Internet Users Group
Minutes of the 14 July 2009 meeting, "Raise Both Hands and Say, Yeah: Multitouch
Technology Is Here"

About 25 people attended tonight's meeting. Josh Zapin facilitated and Jeremy
Kohler recorded the minutes.

----------
MEETING SPONSORS
Microstaff (www.microstaff.com) provides refreshments, Copy Diva
(www.copydiva.com) provides the audio-visual equipment, NCAR (www.ncar.ucar.edu)
provides the facility, and ONEWARE (www.oneware.com) sponsors these minutes.
Thanks to Joe Spinello for recording the podcast.

------------------------------
ANNOUNCEMENTS

Look for a posting on Craig's List and RMIUG for a job opening in a software
consulting/cable software company. Looking for a mainstream Java developer, web
services, agile programming.

There's a good java user group that meets at CU.

------------------------------
INTRODUCTION (Josh Zapin)

Engineers have always been trying to find ways to interact with the human body.
Your car's steering wheel is an example of an elegant interface. Your VCR's
programming interface is probably an example of a bad one. For computers, we
started with punch cards, then came the keyboard and mouse. The original Palm
had a rudimentary touch screen. I think multitouch is very interesting, and
today you see it on the iPhone. Now you can control applications with several
body parts. The iPhone features pinching and expanding gestures to zoom, etc.
Multitouch is a very natural way to interact with a computer. Apple laptops now
have multitouch right on the trackpad as a standard feature. Multitouch adds a
new dimension to the way we do things, and it could open up a whole new world in
the way we interact with computers.

------------------------------
ABOUT THE SPEAKER

Laura Nichols (lnichols@...) is a Senior Technology Lead at Crispin
Porter & Bogusky, a Boulder-based advertising agency whose clients include
Microsoft, Volkswagen, Old Navy, Best Buy, and American Express. Laura spends
most of her time there working on the Microsoft team providing leadership for
the development team, programming, and new technology research. She specializes
in .NET development and has developed a passion for multitouch technology.
Previously, Laura was a Technical Lead at Texture Media, a digital agency that
prided itself on building brands online, where she led the development team for
clients such as Midas, Celestial Seasonings, and National Cinemedia.

------------------------------
LINKS

Natural User Interface Group (www.nuigroup.com)
Perceptive Pixel (www.perceptivepixel.com), Jeff Han's company
22MILES (www.22miles.com), applications for sports
Microsoft Surface (www.microsoft.com/surface)
TacTable (www.tactable.com)
Instructables (www.instructables.com), info for do-it-yourself
Multi-Touch G2 (www.multi-touch-screen.net), software from PQ labs


------------------------------
LAURA NICHOLS

WHAT IS MULTITOUCH?

Multitouch is a set of interaction techniques that allow you to control
graphical interfaces with your hands, using several fingers at once. It's a
system that recognizes more than one input at a time. Examples include the
iPhone, Blackberry Storm, HTC phone, Microsoft Surface, and the HP TouchSmart
computer. There's so much new stuff coming out now that it's hard to keep up.

Companies that specialize in multitouch include Perceptive Pixel, which was
founded by Jeff Han, a pioneer of multitouch.

Another company, called Schematic, is an interactive digital agency. It made a
12 x 5 foot multitouch wall that was displayed at the Cannes Film Festival. The
wall recognized people based on the RFID chips in their nametags, which allowed
the wall to give people very personalized individualized information based on
their registration information. This was entirely developed in Flash.

The Multitouch makes the Cell. These are just boxes that can be scaled into
larger forms by putting the boxes together. It's an interconnecting set of
display cells, arranged in any size and shape. Nobody else is doing this.

Ralph Lauren used its store window to project its web site 24/7. People could
walk up to the window on the street and place orders. It was developed in-house
at Ralph Lauren. A projector puts the web site on the window, and a touch-type
film that was applied to the window senses the touches.

HISTORY OF MULTITOUCH

The iPhone made multitouch mainstream, but it really started in the late 1960's
at the University of Toronto and at the University of Illinois. The Plato IV
computer-assisted education system was built in 1972--researchers used it to
study how touch computers could assist in education. Then in 1982 the Flexible
Machine Interface at the University of Toronto became the first multitouch
interface. In 1983, Myron Krueger developed gestures and demonstrated them using
a light table. He was way ahead of his time. In 1984 Bell Labs created the first
multitouch screen (Myron projected onto a light table, so he wasn't touching a
screen directly). In 1992 IBM and Bell South made Simon, the first smart phone.
In 1998 Fingerworks made some multitouch tablets (and Apple eventually acquired
it to develop the iPhone).

In 2005 Jeff Han came up with some absolutely amazing stuff. He projects an
image onto to a mirror, which bounces to a big screen, and a camera records the
image. He's working on pinching gestures, a virtual typewriter, and is
perfecting pressure sensitivity. You can zoom in and rotate maps, play games,
etc. He has an application that pulls in data from MRI and CT scans, and then
lets doctors drill into the brain and look at the image through multitouch.
Obviously this can be very practical.

Then came the iPhone and, in 2007, Microsoft Surface. It can identify objects
place on a table, like cameras and cell phones. So you can use it to gather
stuff onto a device: Put your phone on the surface, it automatically syncs up
and lets you exchange files.

MULTITOUCH TODAY

Everyone is making multitouch flat panels, and they're getting competitive and
cheaper. There are also multitouch windows. For floor screens, you project onto
a mat and you can dance on it. Could be like the Wii Fit (a pressure-sensitive
platform you stand on to go skiing, etc) but with more capability.

One manufacturer made a screen with real water that runs down its surface--you
touch the water and it refracts the sparks you create, which makes for a neat
demonstration.

How about World of Warcraft? Yup, you touch the screen to move your character,
or hit it to attack someone else's character! (You'd need a really durable
screen for that.)

At the Chicago airport, travelers can use multitouch on public displays to get
maps, weather, etc. that's pertinent to their trip.

REACHING THE MASS MARKET

So you want to develop an application? Ask yourself: Does it work with the
screen you want to use? Is the hardware durable enough to last the expected
lifetime? Is multitouch actually useful and appropriate for my application?

Don't use touch just for the sake of the technology. Start by focusing on the
experience instead.

Making a mobile product? It better be durable because users are going to be
dropping it, etc. And if you're doing desktop-level graphics, you'll need a
powerful CPU.

Don't use multitouch for writing: No one is going to write a novel that way.
Keep in mind that multitouch is for consuming and manipulating data, not
entering it.

Why multitouch applications work: In 1999, the Museum of Modern Art created the
Un-Private House Table. It was a touch table with a hockey puck. It recognized
the data in the puck and let you look at floor plans specified in the puck.

Multitouch is a good medium for sharing: You can push data across the multitouch
table to someone else.

Multitouch is about exploring. Make exploring worth it so you get surprised and
things happen. Design it so you can change stuff and observe the effects. People
should be able to use it to figure out how things work.

Make it simple. It's new, so if it isn't simple, people won't adopt it. A
multitouch table might let you order drinks and play games, but that's it.

The Ibar is an entire bar that is a touch screen, possibly the largest surface
of its kind. It knows when your glass is empty and you need a refill. A lot of
this is just for fun.

A restaurant in Europe lets you order entirely by touch screen: You don't see a
waiter until the food arrives.

Keep it minimal: What you leave out is as important as what you put in. Consider
the Liberty Science Center: It made a cave painting wall where kids could make
handprints, fingerpaint, to simulate cave painting. Nothing more than that.

Size matters: A giant screen isn't always best. The Detroit institute of Art
used a small one to display the Book of the Dead and let people flip through it
to study it--making it huge wouldn't have made sense for that purpose.

Have a clear goal, and make it practical. The Sprint Digital Lounge in the
Sprint store allowed visitors to compare phones on a touch screen without having
to walk all around the store trying to remember everything. You get all the
information in one place. So Sprint had a practical use and a clear goal.

TOUCH TECHNOLOGY TYPES

Resistive: Uses two metallic layers such that pushing makes the layers touch.
It's an old technology that doesn't support multitouch well. For example, you
can't pinch with it. But it's good for single-touch applications. Unfortunately
the layers develop cracks and fail after a lot of use.

Capacitive: This is glass coated with a transparent conductor. Human touch
affects the electric fields and changes capacitance. But it only works with
fingers, not objects. The iPhone uses it. It's cheap and easy to implement.

Projected: Embedded wires register where you touch, and it lets you use objects
like a stylus. It's not mainstream yet. But it's expensive and not workable for
large screens because of latency issues--you can't go much above 42 inches.

Acoustic: Transducers listen to the sound of the touch. This works on on any
surface, so it very versatile. Works even when scratched.

Optical: A projector, a camera, infrared lights, and some acrylic. This is good
for large screens. Infrared light is used for detection: The light passes
through the acrylic panel from the edges and when you touch, you create a dark
spot. The camera senses infrared and sees the spot. The software analyzing the
camera image then registers a touch. Instead of infrared inside the acrylic, can
use lasers that skim along the top of the surface, so you don't even have to
quite touch the surface. Infrared-based systems have trouble dealing with direct
sunlight.

You'll find resistive, capacitive, and projected technologies in notebooks
today.

Keyboard challenges: Keyboards don't work well on a touch screen, so people are
working on adding different types of feedback (which improve typing accuracy).
You can get visual, audio, or tactile feedback. So when you touch it vibrates,
for example, and people make fewer mistakes because you can feel where you are.

Do-it-yourself is not that hard. You need acrylic and a camera that has had its
infrared filter removed and replaced with a visible filter. Use vellum to
eliminate glare. The acrylic must have polished edges so the infrared light can
get through--you can have it fire polished or sit around for an eternity sanding
it.

Tangible User Interface Object is a protocol for decoding detected gestures from
your camera.

The Natural User Interface Group is a great resource for this, and it's very
active.

SUMMARY

Advertising and games are great for getting the multitouch medium out there, but
real-world applications are the important thing. Remember this is for data
consumption, not creation. Hardware is key: It can fail and kill your great
application. And don't use it for technology's sake: Only adopt multitouch if
it's useful.

Now here is the future according to Microsoft: It sees multitouch in file
sharing, medical imaging, sports, ebooks, shopping, credit cards, in-store
navigation and product assistance, engineering, science, and just about every
aspect of everyday life (there's a Microsoft video demonstrating these
concepts).

Q&A

Q: Can I use a "pretend" finger on my iPhone when it's really cold out and I
don't want to take off my glove?
A: Nope, capacitive technology requires a real finger.

Comment from audience: I hear that Apple working on a glove, and there's a
third-party stylus out there.

Q: Is multitouch patented?
A: Apple is trying to patent it parts of it, but the nature of the idea is such
that the whole concept probably isn't patentable. People have patents on the
software of course, but no patents yet on multitouch itself.

Q: Android phones have some pivot hardware, but it's not working well enough yet
for general release. I sure hope Apple doesn't patent that and prevent this
stuff from coming out.
A: I don't think Apple can stop people from doing this stuff. A lot of companies
are out there developing devices and screens.

Q: What about open source software?
A: I like the stuff you can get through the Natural User Interface Group. You
can make it in .NET and WPF (Windows Presentation Foundation). The code is
developed in C++ with a WPF API.

Q: Does tracking more fingers make things difficult?
A: No, performance does not degrade, even with dozens of fingers.

Q: What about adaptive technologies for people who don't have fingers?
A: I haven't seen much of that yet. Obviously, multitouch is not yet accessible
to amputees and blind people. I think there will be way to support these groups
as multitouch becomes more mainstream. There would have to be special
prosthetics that can interact with the technology.

Q: How much computer power do you need?
A: Lots. You need a lot of video memory, and a dual CPU helps. The Community
Core Vision software is a hog.

Q: What's the lowest-price platform out there to develop for?
A: Well there's the HP TouchSmart computer which goes for $1500. That's not
cheap. Then again, my first implementation that I built in the garage was
essentially a cardboard box with a web cam taped to it--that didn't really cost
anything.

Q: So what are you doing with your current garage project?
A: I'm developing applications in my spare time and using the interface for
prototyping. My clients want to see a prototype to demonstrate ideas before
making much of an investment. And it's versatile: It can run on Mac, Windows, or
Linux.

Q: What kind of camera is best?
A: The best deal is the PlayStation 3 Eye camera (you want at least 60 frames
per second, and that's the more affordable camera that can do that). A web cam
will work too, but you have to rip out the infrared filter and replace it with a
visible light filter. Some cameras have the infrared filter painted onto the
lens, so make sure you do your research.