INTERVIEW:
Question 1
What is Jini technology? How does it work?
Answer
Jini technology is technology to allow any intelligent device - a camera, a disc drive, a
refrigerator, a printer - to link together. At the bottom you must have some form of
networking technology. You can connect together cameras, disc drives, computers with
cables or with radio or with infrared. These are all different ways of allowing them to
talk. What Jini is concerned with is what the different devices say. How does a device say
to another one: "I'm a camera and I would like to use you, the printer, to print the
picture I'm taking." Or: "I'd like to use you, the disc drive, to store the
picture I'm taking." Jini allows the devices to talk to each other in this way, just
as people talk, and allows each to offer services to the other. Jini is tiny software that
lives in each intelligent device. Each device, when it connects to the network, uses Jini
technology to look out on the network for other devices, some devices that might want to
use the services that it can offer, some devices that offer services it might want. Jini
allows each device to look out on the network, tell other devices: "Hello, I'm here.
This is what I can do. Would you like me to do something for you?"
Question 2
What can Jini be used for?
Answer
Jini will allow the television set, the CD player, the speakers, the amplifiers, any of
the home electronics devices to talk with each other, to say: "I'm a television set
and I offer a picture and you could use me to change the cable input to a picture, and
then print the picture." So Jini is software that allows any device to offer some
service to some other device. In the kitchen, the refrigerator, the dishwasher, any of the
appliances can offer services; they tell you what it is they do. That would enable you to
control them, because a major step forward for any city is to allow homes to control their
appliances inside the home to bring down the amount of electricity they use. Today, on a
hot day throughout Rome air conditioners, refrigerators, all turn on, which requires ENEL
to generate lots of electricity. If we could save a little bit of electric power, bring
the peak load down by 5% or 10%, we could save tens or hundreds of millions of dollars of
generation capacity. Allowing devices to speak to each other can cause tremendous savings
for the society as a whole. Jini is just software. It allows a smart device to communicate
with another device. Here's an example: a cell phone is a computer with a radio, a
keyboard and a screen. This device can send across the network a piece of software saying:
"If you would like to display something on this screen, use the software, and it will
come back and display on the screen." So if I'm building a messaging service, I might
like this Jini-connected device to offer that display service to me. If I'm using my
television set at home and I'd like to control it, the television set will send its
software that makes it work using Jini out so I could talk to it with my cell phone, so I
could control the television with the cell phone. Allowing these devices, which all have
computers in them, to talk to each other let's us redesign what we use in our homes to
listen to, to look at, to cool things, to heat things. We can begin to let them all create
new kinds of appliances. And this, of course, allows new businesses to begin. Jini was
developed by SUN in cooperation with hundreds of other companies whose suggestions about
what they need were put together into this very simple design. In the last few weeks, 37
companies, including Motorola and Sony, announced on one day that they would use Jini in
their devices.
Question 3
Who created Jini?
Answer
Jini was developed by SUN because we knew all of our partners. The cellular telephone
companies, the disc drive companies, the printer companies, all of them have the same
problem. When they connect their devices to another device, they want the other device to
know how to use them. For example, if you have a PC and you want to connect a printer to
it, the software to make the printer work must already be in the PC. This is backwards,
it's not easy. The disc drive or the printer company wants to put the software in the
printer or the disc drive. So when you connect it, the software comes from the printer or
from the disc drive over to the computer that will use it. That way the software is never
out of date. So all of the companies that make electronic devices run by software wanted
to be able to keep the software in their device and connect to the network, move the
software to those that would use the device and always be up to date. We recently
announced that 37 companies would support Jini, including all the consumer electronics
companies - Sony, Philips, Thomson, Mitsubishi, Sharp, companies that make television
sets, companies that make CD players, companies that make devices inside our homes, all
the printer companies, Hewlett Packard, Epson, Canon, all of the cellular telephone
companies, Nokia, Erikson, Motorola - all these companies want their devices to be able to
work together with each other. With Jini the software in each device moves across the
network and joins with software from other devices to form a community. All of a sudden, a
television camera can find software coming from a printer and send its picture to the
printer or find software coming from a disc drive and send its picture to be stored on the
disc drive. We call this "spontaneous community", a community formed of all
these different hardware devices that do very different things able now to use each
other's capabilities. The reason this is so important is that it reflects change in
computing from the old style, where we think about everything on a disc in one place,
waiting for a device to become connected. I have here a PC. In this PC, sitting on my disc
and this one has a 4GB disc, I have megabytes and megabytes of software for disc drives,
for printers, for modems, for devices I will never use. The software is here waiting. So
if I bring a printer from the store, a new printer and plug it in, the software is here,
already waiting for the printer. Microsoft calls this "plug and play". That's a
good idea. When you buy a new printer, you should not worry about how to make it work. But
the problem with this centralized model is that you must have the software here before you
buy the printer. In the new world that will not work. In the new world with devices
becoming smarter, with chips in everything, when you buy a new disc drive or a new
printer, the software will be in the printer not here in the computer. And when you bring
your new disc drive or printer home to connect to the computer, the software will come
from the printer into the computer, so it will just work. That's quite a different model
from the older model Microsoft has. And that's the model for the future. These PCs, there
might be 200 million of them. The devices in cell phones, in switches, in beepers, in
printers, there are billions of them, thousands of millions of them, far more than
personal computers. So when we talk about the devices that people will have, people will
have wrist watches, cell phones, beepers and dozens of different devices that all can
communicate together using Jini. So our plan is to enable any manufacturer of anything, a
cell phone or a camera, to move software from one to the other. How might that work? I
might be able to move the software using Jini from the camera to my cell phone so I could
control the camera from my cell phone. With Jini that's easy to do. So the future allows
us to take devices that are independent today and together make something new that never
existed before. That's the power of Jini and that's why all the computer companies, all
the consumer electronics companies, all the printer companies, all the disc drive
companies have said that they will agree and standardize together.
Question 4
You talked about sophisticated, hard-tech devices, cellular phones, beepers, so on and so
forth. Nicholas Negroponte, who I am sure you are all familiar with, has indicated the
fact that such developments could even go all the way to involving let's say everyday
objects, so-called things that think: tables, tea cups, coffee machines so on and so
forth. What is your view on that?
Answer
If you look at the way computers have developed, in the last 40 years the computers have
been big. The mainframe computers were big boards with many chips on them. That changed in
1982, 1983 when microprocessors came out. These powerful chips let us take many of the
separate chips on the board and turn them into one chip. And this world of microprocessors
and single board computers, the Apple II, the PC, the SUN workstation, these single board
computers have dominated the world of computing for the last 15 years. And today we're at
the edge of the next big change, the move from the single board computer to the single
chip computer. Chips cost very little. The first one costs a billion dollars and the next
one costs nothing. When you're thinking of making a chip, the design part is expensive,
building the plant to fabricate it is expensive but the chip itself costs nothing. So
there will be thousands of millions of chips. So that's computational power, the ability
to be making decisions, that's what the chip does. The other part of it is the sensor.
That's what measures things. And today using the same technology we used to make chips, we
can make sensors. How do you do that? The chip is a small piece of silicon. We can make a
small swimming pool in silicon the size of a millionth of a meter across and leave a
little diving board hanging over the hole we dig in the silicon. Then that measures
whether or not you move the chip. Because if you move it, the diving board vibrates. So
you can make an accelerometer one millionth of a meter across that costs nothing. So we
can put those in devices that measure impact. You can put it in a letter so you will know
when you receive a letter if the letter had been dropped. You can put it in a coffee cup
so that you can measure how strongly the coffee cup banged the table when you put it down.
We can begin to take the information in the world around us and use the sensor information
together with the decision-making capability of the chips and both of these components
cost nothing. You can take a piece of fiber optic cable, just glass, and put a chemical on
the end that only reacts with certain molecular shapes. That you can put in your
bloodstream to detect sugars. Today at the pharmacy you can buy a device of US$50 that
measures the amount of sugar in your bloodstream and you need to if you are a diabetic.
There are sensors which measure electric fields: you could put a sensor on your heart that
would predict a possible heart attack. The electric patterns of a heart attack are very
well known. That would then communicate with an alarm or possibly with something your
mother would buy. A small voice that would come from your wrist watch and tell you not to
drink that cup of coffee because your heart was already going too fast. So the ability to
measure inside the body, the ability to measure the state of your body, gives us new
medical diagnostic tools. That's all made possible by small chips and small sensors. How
do they talk one to the other?. Jini allows small sensors, small chips, to talk each to
the other in a very simple way. What we're building is something new. We call it
"distributed computing". Everything comes alive and all the devices that have
something to say: "I'm hot" , "The heart beat is going too fast" ,
"The air contains chemicals that are not good for the human being" , "The
water in the river is becoming too hot from industrial outflow." All of this
capability to measure and see makes things that are invisible to us now visible, so we can
start to make decisions about the world around us, the world outside of us and the world
inside of us.
Question 5
The real world and the digital world are merging and becoming interrelated. This is
changing our lives right now and we can expect this to come even faster and more dramatic
in the future. But this process seems to be driven by a handful of big names, Microsoft,
to start with and also relatively new groups like SUN, AOL and Netscape. Do you think it
could be dangerous if such a far-reaching process which affects so many people continues
to be driven by few entities?
Answer
Those that are driving this process of digital communication are the telephone companies.
Long before SUN or Microsoft or any of the small companies that came about because of the
microprocessor, the telegraph companies sent digital signals so that human beings could
communicate. Then the telephone companies that first began sending voltages down the wire
found it was hard over a long distance to make it sound good, so they converted to making
the signals digital. That's just ones and zeros. And if you could look inside the phone
wire, you would see all that happens in transmitting a signal is that there's a voltage
that goes up and then a voltage that comes down and that's a one. Then there's no voltage,
no voltage and a one again. So you can send a signal that simply looks like a square wave,
and you send it down the wire. When it becomes noisy, because static from elevator motors
and the spark plugs in cars make spikes in it, you can pass it through a tiny computer
that gets rid of all the noise. When it sees an ugly signal come in, it changes it into a
perfect square wave again. All that digital means is perfect copy. Now, in the world
everything is analogue. I have a different sensation as I press harder and harder. That's
not digital. The only point of digital is to allow us to take information and transmit it
from one place to the other and at the other end have a perfect copy of what we started
with. This is good if you like music, because you can transmit voice or sound or pictures
perfectly anywhere in the world. All of this movement was brought by the communications
companies. When we began to bring that capability into a device that you could look at,
with a computer inside it, Xerox, the companies that had invented the personal computer,
that made something a human being could use, gave the technology away free. And when
people built systems that linked many computers together, they used Ethernet, a digital
standard for moving packets of information among computers, used in every computer. That's
a free standard that everyone uses. On the Internet, TCIP, a packet standard that anybody
can use that allows you to put information in a little packet and throw it out onto the
web of interconnected machines, and it will find a destination no matter what pathway it
needs to follow. So these innovations that brought about what we loosely call the digital
world are really innovations in moving information from one place to the other with no
change, perfect sound, telephone from Rome to Beijing and it sounds as if you're in the
next room. That capability lets us now link everything else, all of the databases, all of
the sensors, everything together. The goal in all of this is to allow people to build
something new, depending on what they want to do with no one controlling it. But you can't
have everyone doing something different. So bit by bit we have established standards in
linking the computers together, in linking the phones together, in linking the television
cameras together, in linking the recordings both on CD and on the Internet together. Soon,
we will just think of all the interconnected digital universe as a giant source of
information. Now, the next challenge is to allow people to learn to read and write,
explore and find things. In each of these areas big companies spring up because they're
able to do something better than someone else. Microsoft has emerged in PCs as a company
that has taken technology usually invented by someone else - in fact I think it's safe to
say Microsoft has taken technology always invented by someone else - and managed to put it
into small PCs for 100 or 200 million people. The companies that have made a much deeper
impact are the consumer electronics companies and the telephone companies that have put
devices into the hands and into the ears of hundreds of millions of people. There are 800
million telephone lines in the world. You can see how far we still need to go because
there are 6 billion people and only 700 or 800 million telephone lines. So most people
still are not in the conversation. The challenge for all of us now is to extend this to
everyone in the world, so that a child in the most remote village has a sense of being in
a conversation with other children able to read, to write, to exchange ideas with anyone.
That comes about as the devices that make this possible become cheaper and cheaper and the
networks that make this possible become ubiquitous, providing a pathway for everyone to
talk with everyone else. Now I didn't get to the industrial structure. Microsoft has very
little to do with all of this when you get down to it. It's just our view of it that
Microsoft never mattered. Here's a question to you guys. It's a test. How big is
Microsoft? How many people work there? Answer: 26,000 or 25,000. How many people work at
SUN? 27,000. How much money did Microsoft make last year? US$14 billion. How much does SUN
make? US$12 billion. So we're almost exactly the same size. The same revenues. Now, how
much is the capital valuation of Microsoft? US$200 or $300 billion? Ours is US$60 billion
or something. So there's a big difference in the valuation and that's why Bill Gates is
rich. But the technical impact of Microsoft on things that matter is only the impact of
widespread mediocre software. And the problem for us didn't exist a few years ago when
Microsoft didn't have very many computers that it affected. The things that are important
for us are the computers that run the banks, the trading floors, the air traffic control
systems, the electric power grid, the computers that run the cameras, the watches, the
cell phones, the devices we use. Microsoft is not present in any of those places. So, it's
often easy, because people use PCs, to think that they're an important device. They're
just an accident for the moment. And the inevitable change in the way the world works is
to think about the computer as a contrast between the PC and the cell phone. Let me show
you what I mean by this. This is a PC, an elegant manufactured and designed device, an
elegant result of a mediocre design. I have a huge disc in this, loaded with software I'll
never use, software waiting for a disc that I'll never connect or a modem that I'll never
connect or some printer I'll never connect. And the problem at the basis of this is the
problem of every computer of the last 40 years. When you start it, it first looks for a
disc. It finds a file on the disc with a specific name and it starts executing the
instructions from the disc. Let me show you the modern computer. When you turn this cell
phone on, it tests itself and instantly goes to the network. There is no disc here and
there's no disc in the billions of computing devices that exist today. Today's automobiles
have 100 to 150 processors in them, when you start them, they don't go to disc. They look
on the network for any information about what to do. So we're moving from an old world of
disc and file-based computing to a new world that's network-based.
|
 |
