Scott McNealy
Thursday, April 3, 1997
John Gage: Good Morning. Well, everyone looks awake. You shouldn't be if you stayed up `til midnight last night meeting people. First test; how many of you met somebody new yesterday? [hands go up ] Pretty good. But some of you are still failing the shyness test. So remember, don't be shy. Whatever your national or cultural background might tell you is the polite thing to do, ignore it. Find somebody that looks interesting, try to read their name tag. It's okay, because our name tags have things written so small, it's okay to bend forward and look at someone's chest. It's okay between now and Friday evening, it's okay. Find out where people are from and say hello to them. Don't be shycompanies are forming right here.
Last night the volunteer crew made some ribbons: authors, come get your ribbons after the keynote. We've had them signed. Some of them, you'll notice, will have a little scrawl on them. That's Scott McNealy's son. How old is your son, Scott? 16 months. So he tried to write ``Java'' but it's not quite legible. So you'll get some of Maverick's signature on that.
What other announcements do I have? Oh yes, most of what goes on in each session, transcripts of the keynotes, slides, and audio are going up on the web site each day. It just means, there goes another weekend, reading all that material for all those sessions you missed.
Each day as a session finishes, the real time news organization, JavaWorld Magazine, will be putting up summaries of what happened; along with pictures and clips. Many of you are being interviewed about what you think went on, and you'll find that at http://java.sun.com/javaone It's the JavaOneSM momentary newspaper for this entire event. So you can find these stories if you have a moment.
We're also going to put out CDs after this thing is over, with everything on them. So if you're in a country that doesn't have tremendous bandwidth, then you can get it all on a high bandwidth in a small package through the mail.
Now we've made up a lot of the ribbons for Java authors. I think we have probably 300 here, so we've made up enough ribbons. We couldn't get Bill Joy to sign all of them, but we've got most of them signed by someone, by Alan, or by me, or by a number of people.
And then, following Gates' performance last night where he made a joke of security, which we thought wasn't really in good taste, we made up a lot of the ActiveXTM bug bagger ribbons. So if you've found an ActiveX bug, fallen into it, or have noticed that your bank account has been declining because the of the chaos club.... You can't stop these ActiveX things from addingnot just data, but executables at the end of a random file. You can't stop it, we have a serious technical problem at the bottom.
It sort of reminds you of, ``What is DOS?'' ``Is it a system?'' No, it's just some software. People wrote things. People threw them out. People use them. People had to fix them, and people wasted more timewhen you think of just the amount of engineering time wasted in the 640K problem alone. I'm temptedand I think I'll do itto put up a web page called ``Billing Bill.'' Everybody comes to the page and just says how many timeslet's use a unit, how about minutes, maybe hourshow many hours have you wasted in your professional life using something from Microsoft that crashed; that restricted you to 640K; that could not be relied uponhow many hours have you wasted just for the damn things to boot up? How many engineering hours?
We'll work on this idea, but there's a metric about how much of your time is wasted because of bad engineering, because of fundamentally bad design. So we'll work on that.
Now we have a new ribbon category, which was suggested yesterday; it's a category for Java teachers. So if you're a teacher of Java, we've got a ribbon for you. I'll just show you who you're going to look around on the floor to honor. The authors' ribbons are a bright lime green. The VM porters' ribbons are a blood red for blood, sweat and tears. Anybody wearing this one did the serious work of putting a VM on some architecture. The teachers' ribbons are blue, a sedate color. The people that found any kind of a problem in Java security, any flaw in a VM, are going to wear this bright pink. So you can ask them how they found it, what it was, does it exist, was it anything fundamental (as we believe about ActiveX), or was it something that was passing, that was a bug, that was a mistake in implementation, someone that didn't read the spec, which unfortunately occurred. And then here are the bright yellow ribbons for the ActiveX bug baggers that have stumbled in a hole, and found a bug in ActiveX. We expect there will be a lot of people wearing these.
All right, so that's the ribbon story. Then this afternoon, when Dana Carvey presents, there is a request from Dana Carvey's theatrical agency that nobody bring video cameras or cameras to record the performance. This is just how things are in Hollywood.
I think with that I've finished what I need to say. Is everybody moderately exhausted so far? If you're rested, you're not working hard enough. And now I'm going to have someone come up to tell you exactly how much harder you're going to have to work, just what the standards are for moving all of us forward together, every company forward together, Scott McNealy, the President of Sun Microsystems.
S. McNealy: Thanks. It's nice to be here and it's nice to see you all here. Thank you all for getting up earlypre-programmer hours here. I've got a lot I want to cover. I've probably got to do my obligatory bashing. I want to talk a little bit about where the world's going. I want to do some demos, and we'll get to the top ten list. That's another topic. By the way, cameras and recording equipment are encouraged here. [laughter] And contrary to popular belief, John Gage does not have a very large home in Rancho Santa Fe. He just looks that way.[ laughter]
So what do I want to get done here? I guess I'd like to reflect for just a couple of seconds. The world's really changed in the last couple of years. If you think back two years ago, there were a lot of computer companies who were actually in charge of interfaces. But the Web's in charge now. I'll talk a little bit more about that. We didn't have JavaOne. Java was just another cup of coffee back then. Windows was the number one unit-volume platform in the industry. That's how long ago two years was. Not the JVM. And Microsoft was giving the keynotes in the main ballroom during the day, not late at night in some breakout room in some second budget trade show.
My mom's over in the corner, and she told me, ``Now you be nice to Bill today.'' Yes, Mom. Oh, I can't. I can't. I'm accused of personal attacks. They're not personal attacks, but they are attacks. But who am I going to pick on, DEC? That wouldn't be any fun. I wouldn't know what to do. I don't know how to fill a whole keynote with content. [laughter ]
I was going to do my top ten list, and I know some of you enjoy them. Some of you don't, but that's tough. But you know what, two years ago nobody copied my top ten list except Letterman. Microsoft did one last night. So I don't think I want to do one today. [audience groans ] It just seems too redundant or copycattish or.... You know, it's kind of funny. We were going to give out ribbons for ``I blew off Bill last night,'' but there wasn't enough ribbon in the area. [laughter] We figured that would be redundant with your JavaOne badge, so we aren't giving those out.
But he did a top ten last night, only it was little different. If I'm going to do a top ten, I'm not going to hire some marketing person to go give it for me. He actually had some marketing guy give it for him! You were there? He shouldn't have done that. He actually wouldn't give it! He had some marketing guy give it. The topic the marketing person used was, ``Why I don't want to be the ActiveX security spokesperson.'' There were more than ten reasons, let me tell you. Anyhow, I thought, ``What could I do different?''
So it was about 9:00 p.m., and I heard that he did a top ten list, so we put together a different list. Clinton hurt his knee and a ticket became available for the Academy Awards last week, so I went to the Academy Awards. I thought, ``If we were to run a motion picture Academy Award ceremony, what would that look like? If I was the Academy, what awards would I give out.'' So I thought I would walk through the ``Scott McNealy Academy Awards,'' if I were giving them out and if we had awards. You have to kind of assume these are all movies.
So the best horror flick? Microsoft Security Model.
How about the best tragedy? Apple. Oooh, low blow.
How about best romantic tragedy? Larry Ellison does Apple.
Best science fiction? Your application doing long division on a Pentium.
Here we go. Best fictional screen play? Microsoft's Internet strategy.
Now we're going to get into the hardware; I know it's a software thing but I've got to do some hardware here. Best sequel? DEC is back, part 48.
Here's another one. Best short story? The future of Hewlett-Packard's Precision Architecture. Ooooh. Hey, deal with it! This is going to be in the press, so we've got to get in on the hardware. So you just have to deal with these.
Posthumous lifetime achievement award? SGI. Ooooh. It was late.
Best one-liner adaptation from Jerry McGuire, ``Show me the money.'' Microsoft for, ``Give me your money.''
The best cartoon? Bob. Or is that CE, I can't remember?
Anyhow, and the last award is for best screenplay, Java by James Gosling. Thank you.
So now I suppose I've got to get into content, as you can see. These are the notes from my marketing people. ``Be reserved.'' [He shows the audience a printed seat card that reads, ``reserved.''] It's on the other side. I can't read it. [He flips the card over.] This is my Steve Jobs pose. [He puts his hands together in a prayerful state.]
I want to talk about a few things, how we're trying to handle Java. A lot of people are talking about hijacking Java. I want to talk about that. I want to talk about who's in control. I want to talk a little bit about the scaleability thing, because I really do think that opens some doors.
The first question a lot of people ask me is, ``How you going to prevent Java from being like UNIX?'' The first thing I say to them is, ``Hurt me'' if Java hurts Sun like UNIX hurt Sun. I hope Java is the same kind of disaster that UNIX has been for us. We've gone from nothing to a 20,000-person company, grown, and made money every year. Please be that kind of disaster over the next 15 years, right?
But we did try to learn some things, and unfortunately, or fortunately, depending on how you look at it, we feel a fairly strong sense of duty here in terms of our stewardship with Java. We're trying to do it in a different way. We're trying to do it in a different way in that we weren't in charge of UNIX. AT&T was in charge of UNIX, for some of you who may remember, and I don't think that was necessarily their business or that they were very focused on it. It probably wasn't handled like we would have liked it to be handled.
But Sun has been committed for 15 years to published open interfaces. It's kind of weird, kind of strange, and a lot of people have looked at us as if to say, ``Well, how are you going to lock them in? How you going to own them, how you going to control them?'' I guess I just grew up with the automobile industry, where Ford didn't own the fact that the brake pedal went to the left of the accelerator. That just wasn't their patent. You didn't patent where the brake pedal went. You didn't patent interfaces. You tried to have all of the bumpers at the same height, so that when you had a wreck, you didn't have bumper to jaw.... that just didn't seem to be the right interface.
I got into the computer industry and everybody wanted to own their own written and spoken language. We have decided to take a very different approach for the last 15 years, and that iswe're going to allow interfaces to be open. Now, we were pushing uphill, and all of a sudden the Web happened. And all of a sudden the Web said, ``We like that strategy. We're going to embrace that strategy and we're going to go with open interfaces.''
But, you can't necessarily use the political process effectively to make these standards move forward, and the de facto process tends to work a lot better. But also, collaboration tends to work a lot better. People say, ``There's no way you at Sun can keep up with the R&D at Microsoft.'' I reply, ``That's actually pretty easy, because our R&D is our R&D, plus a lot of Microsoft's R&D, a lot of IBM's R&D, a lot of Netscape's R&D, Oracle, Sybase, etc.'' All these folks are all putting energy into this architecture. When you open up the interface, you stand on the shoulders of everybody else's R&D. So our commitment in terms of stewardship for Java, is to take the best we can find from anywhere in the industry and put it in the JDK.
We have some rules. First, you can't charge us for it because we're going to give it right back out through our licensing mechanism without raising the price. And second, we're going to require that it be intellectually free so that we can put it in there without getting sued for using that technology. And that's a very, very different strategy.
So JDK 1.1 was not written by Sun exclusively. Very significant portions of that technology came from our partners. That's unusual; that's different. And it's a very, very different mechanism. It's a very different stewardship. The interfaces are open. Just go down to the bookstores and if you want to read 825 pages of James Gosling, Guy Steele, and Bill Joy, pick up The Java Language Specification, it's all therethe grammar, the syntax, the alphabet characters. There are no secrets, no hidden agendas in the language. It's all there. That's our strategy. That's our focus.
So I just wanted to make that pointthat we're trying to operate very differently from the UNIX space, but still take advantage of the openness, the choice that the UNIX space offered. Anybody that has any comments or suggestions, feel free to send them to us, and let us know. We're trying to refine the process and make sure it's moving forward. The last thing we want to have happen is a process that coagulates and moves slowly. And if anything, we've surprised people at how far and how fast we've been able to move in the first 700 days from launching Java to having the kind of programming environment that you all are dealing with today.
The next question that everybody wants to talk about is, ``Is somebody going to hijack Java?'' And the two big worries are: will Microsoft hijack it, or will Sun hijack it? I think at some point you have to ask yourself, ``Can Xerox hijack Ethernet?'' No, absolutely not. Xerox invented Ethernet. A lot of people don't know that, but they did. Can Paul Allair stand up and say, ``We've changed Ethernet and I want you all to send me a big bag of money because I changed it?'' No, at some point you just can't and it actually is a much bigger and broader process for people to own this.
So I don't feel like we can just hijack Java, take it, and run with it. The interfaces are open. People will engineer us away. IBM and Oracle will get together if we do anything wrong, and push us aside. We are under very strong pressures from our other partners, from you all, to go do the right things, and that's not necessarily going to allow us to create a $100 billion market cap for Sun Microsystems in the short term, but that's not necessary either. So we think the strategy works. We think we can find a way to be quite successful and quite profitable without having to own the language. We'll just make money doing things in the language and spend a few extra dollars stewarding the language, if you will, forward into the next century.
Microsoft, can they hijack it? No. Clear and simple. Either it's Java or it's not. Either it passes the Java compatible test suites or it doesn't. Either it's 100% Pure Java as an applet or it's not. And if it does, it's Java. If it doesn't, it's not Java. Now you all have a choice. If it's not Java, it's probably Windows, right? Because they're the only ones who are actually, not necessarilyI mean, I love it. They kind of say, ``We're 100% committed, but we're not 100% pure.'' My wife is over in the corner. I don't think she'd buy it if I said, ``Honey, I'm really committed, but I'm not 100% pure.'' I just don't think that would fly. [ laughter]
There's no way they can hijack Java. They can continue to push Windows with some Java code in it. You all have three choices. Write in Windows only. Write in Java only. Or write in both. But there's no way they can hijack Java and just change Java or be Java. And you all have to make the decision. Do you want to run cross-platform? Do you want to run without viruses? Do you want to be part of what we're all doing here? Or do you want to write to the Windows environment? That's the choice.
I just don't even know how you'd hijack it. It's like saying LAN Manager could hijack TCP/IP at this point. I mean, LAN Manager is there. It exists. It's just a matter of what kind of unit volume it has. Java is here to stay and I don't believe anyone, including Sun, can steal it. And that's because the Web is in control now.
We've been watching over the last 15 years how the proprietary written and spoken languages of computers have gone away one by one: Wang and DEC and Apple and IBM and all these companies have basically hit the wall because the world will no longer buy their lower-volume, proprietary environments. The Web is now in charge. What I tell every enterprise customer that I see is, ``Implement what the Web is doing, because the Web is where the energy is. The Web is where the applications are. The Web is where the developers are. The Web is where the content is.''
And if you don't do what the Web is doing.... and what is the Web? The Web is the customer. For the first time the customer is in charge of our industry. That bums some executives out in our industry. We've actually been trying to do that for a long, long time. That sounds a little self-serving, but it's kind of fun now to actually have our strategy be understood and appreciated because we've taken a few lumps over the last 15 years for trying to put the customer in charge, for giving them freedom of choice, as opposed to freedom from choice.
So anyhow, that's what's going on. That's our view of this thing, and I think it's a pretty interesting step forward in the industry. Before we get into some demos, I'd like to just talk a little about the scaleability. Alan talked about the smartcard, the supercomputer, and scaleability. I really love the stuff we're doing on the enterprise side, because that sells lots of big Sun servers and that's cool. [laughter] And the other thing is that every time there's an IP client put on the networkwe call them Ncs, I call them IP clients or Java clientsyou need some percentage of a server. So the server business is going to be good.
But what's going to drive all that is the ultimate thin client, which is the Java card. [holds up a smartcard] Now, for those of you here in the U.S., we haven't really gotten it yet with smartcards. I mean, you all have one in your set-top box maybe. But we don't carry them around. Gemplus is shipping a million smartcards a day, including weekends.
Now we're talking to the silicon makers. We've got 85% of the smartcard market. We're actually talking to the silicon makers to put Java acceleration on the actual silicon. Now this world has evolved in a pretty interesting way. I've got a Swatch here, if I can get the box open. This is, I don't know if you can see here, this is a clear Swatch. And there's a little blue antenna ring around the dial. This is a real thing. And there's a little smartcard chip in there, which runs it without a battery. And you wear this watch in Switzerland, and when you go skiing, you go up to the turnstile, and there's a little proximity reader; not a contact reader, but a proximity reader that can transmit and receive to and from your watch, even though there's no battery on your watch for the antenna. And it actually talks to the smartchip in your watch and debits your watch one lift ticket, and opens the turnstile. Now is that far out?
I just wanted to show you that technology so you can think about what is possibletalk about a software opportunity. Now I'll walk you through a George Jetson kind of day. You get up in the morning, you go to your car and do this to the door [He pretends to press his back pocket against the side of a car.] and the proximity reader opens your car door. You stick your card in the dashboard of your automobile. You type in your password. So this is your key [holds up card], but you need the password with it.
And then you drive to the office. This opens the door into your building. You put this in your Java Station, and you type in your password and it takes you right to your web page, right to your webtop, and it gives you access to all your applications, all the rest of it, so again you have double security in that environment.
You go to the airport, you put this in the Southwest kiosk. It tells you what your seat is and prints out your boarding pass. You stick this in the back of your airline seat, and AT&T beams down your phone number right to that seat, because of the information stored on your smartcard. You get to the hotel, you don't check in. You go to the lobby. You stick the card in a kiosk. It secretly shows you your room number. You go right up to your room number, stick this in the door, and it opens. You stick this in your set-top box. TCI beams down all your favorite shows from your home town right to your set-top box and bills you for all of that, and it's all taken care of.
And you can use the Java electronic commerce APIs with things like the affinity APIs so you can do frequent flyersyou can get miles for everything you did, right? And we can see that you're going to have a wallet full of smartcards that get you into places. Federal Express will give you a smartcard. Your bank will give you a smartcard. Safeway will give you a smartcard. You name it, you'll have a smartcard for every environment. That's the opportunity. And you can do it safely and securely. I mean, this is what I call the ultimate Java Station, right? [holds up a smartphone ] It's got a keyboard, a display, a speaker microphone, network, port, the whole deal. That's the market, that's the scaleability. Other than security, which is one of their Achilles legs. I believe the thin client here is the other Achilles leg that the Captive X world really is going to have to deal with. So anyhow, enough of that.
I thought we'd bring Miko back up on stage. We heard he was very well received yesterday, so we thought we'd have him up on stage to do another demo. Take it away, big guy.
Miko: Thanks, Scott. Scott just talked about the openness of the platform, to the extent that we would find the best-of-breed technologies available anywhere and allow them to come in and become part of this Java family. Well, we've got a very interesting story for you today, and I'm going to share it as the first demo.
Sun was looking for a partner for JavaSound. Now, you've probably seen Duke. You probably heard the ``whoooppeee'' and the space music.au. Well, we wanted to do a slightly better job with JavaSound.
So we searched high and low to find a technology that was fitting to be called JavaSound., and we ran across a company in Silicon Valley called HeadSpace. Now HeadSpace is a very interesting company with some amazing technology that we're going to show you in a moment. What's interesting about this company is that they have an interesting CEO. The CEO of this company was actually a pioneer of the new media in the 1980s, MTV, and is now a pioneer of the new media of the 1990s. I'd like to introduce you to Thomas Dolby Robertson.
T. Robertson: Thank you very much, Miko. I have to say, apart from the fact that usually at this time of day I'm just going to bed, I feel right at home here, because I look out and I see this sea of white male faces, and it's just like playing a heavy metal concert. [laughter ]
You may be asking yourselves, ``What business does a pop star have going to Silicon Valley and licensing code to Sun Microsystems?'' Well, to explain this in my terms, which are human terms rather than programmatic terms, I have to back up a couple of years.
A couple of years ago I got really interested in the Internet. And I started looking at the way that sound and music were being dealt with on the Internet. And I thought, ``There is something wrong with this picture,'' because when I look at the way text and graphics and information are being done on the Internet, it's a whole different model from the old way where you just took a big fat linear object and distributed it to a wide range of people. And yet what I saw with sound on the Internet was that people were taking their digital master recordings and trying to pipe them across a telephone line. And I feel this is entirely the wrong metaphor. Let me try to explain what I mean.
This is an 1885 volume of Shakespeare's Hamlet [holds up an antique book]. It's very beautiful. But it's also rare, and it's perishable, and it's a little bit too heavy for me to send to school in my kid's backpack. This is a Reader's Digest version of Shakespeare's Hamlet [ holds up a thin paper booklet], and it's affordable enough to put on every newsstand, but it's what you might call a lossy comprehension technology inasmuch as I don't think the author of Hamlet would have been too pleased about the way his work was represented in this edition.
On the other hand, this floppy disk [holds up a diskette ] is a hypertext version of Shakespeare's Hamlet, and it's really cool, because every word is exactly the way that Shakespeare wrote it, and with a single command, I can reformat it, change the front, and the size. I can search it, I can hyperlink off to another place on the Net, and do a cross reference. This is why hypertext is so cool, way cool, and this is why the Internet is built around hypertext.
Well, how does this analogy work for sound and music? Forty minutes of great music; rare, expensive, and perishable, and not very easy to send over a telephone line [holds up an old record]. Forty minutes of real audio [holds up a real audio box]; is very easy to send over a telephone line, but a kind of a lossy technology and one that, as an artist, I'm not very happy abouthow my work is represented if the output to the audience is not something where the fidelity is guaranteed to me.
Forty minutes of hypermusic [holds up a diskette]cool concept, very small file sizes, very flexible, easy to distribute over the Internet to a wide range of people, and yet malleable enough that at one time, you can do all sorts of different things with it.
So hypermusic was a concept that I just got fixated on a couple of years ago. And I went and talked to a few high tech companies, but they didn't want to know. I thought to myself, ``The only way that this is going to happen is if I do it myself.'' So together with some really excellent software engineers, I formed HeadSpace a couple of years ago, and we set about building what we intended to be the very best software audio engine anywhere. I wanted this to be something that could not be owned, and yet I wanted this to be something that would be everywhere. I wanted to paint the world with this engine.
So I had a couple of choices of people to go and talk to. And during the period that this was unfolding, I was thinking to myself, ``This is good.'' I looked around at what other compatible technologies I could see in the world, and it didn't take me long to figure out that the right place for us to be was in Java.
So in the fall of 1996, I went to talk to JavaSoft, and I played them some examples of technology and they were instantly very impressed with the sound quality. And we talked things through and at great length we arrived at a point where a deal was struck to integrate HeadSpace's technology into Java, and as you can imagine, we were pretty jazzed about that.
But they did have some very stringent requirements for us, and they said, ``In order for your technology to be compatible with Java, you're going to have to show us that this is a head-to-head match.'' Now Sun said, ``We don't know anything about audio software, but we've had a look at a bunch of different technologies that are out there, and frankly we don't see anything that is really worthy of being in Java.'' And I said, ``Well, we've got something to talk about, because I don't like what's out there, either, but take a look at this.''
So they set out the requirements they had for the appropriate audio technology to go into the core of Java. Now these bullets [on slide ] are ones that any of you familiar with Java will instantly recognize as the tenets of the Java creed. So I don't have to preach that to you again. What I did set out to do was to show Sun how HeadSpace's technology was a direct head-to-head match.
Well, in the first instance, it needed to be compact and it needed to be secure. It was certainly compact. One of the most important things was to make it very quick to download sound over the Internet, whatever the bandwidth. And that's pretty much just a numbers gamethere's nothing particularly revolutionary about that. You just want to get stuff from point A to point B as fast as you can.
They also said that they needed it to be secure. Now as a Java Foundation Class, we would be guaranteed a level of security, but I had a certain aspect of my own that I felt, as an artist, I needed to add to this, and that was that there was a lot of intellectual property embodied in the work that I put out. And there was a general perception that at the moment, as musical sound got onto the Internet, it was public domain. And that just didn't work for me. So I insisted that there was a capability to make this media have a digital signature, or watermark, if it was going to go out on the Internet.
To demonstrate that, the file that Miko played when I came on stage was a recording in a format called rich music format. Now one of the neat things about this is that I can embed my copyright information into every copy of this file. And this copyright information can be displayed by any user anywhere the file is heard. So this lets you know, not only who the author is, but also how to contact the author or publisher and what kind of licensing deal was made.
Sun also insisted that the technology be completely platform-independent. Well, HeadSpace's technology runs on six platforms today, and the code base was designed to be very, very easily portable, to the extent that when we did our port to Solaris at Sun's request, it came up in about 10 hours, and that kind of spooked me, frankly. But what's neat about it is that I can author my file once and on any of the platforms that HeadSpace supports, it will sound exactly the same. Now if that sounds familiar to you, it's basically the same thing that's really important to Sun about cross-platform compatibility.
The technology also needed to be familiar and open. I wasn't going to insist that any of my peers abandon the tools they've been using for a decade or so, so we embraced popular file formats such as MIDI, WAVE, AU, AIFF, and MOD. And we encourage audio professionals to keep working with the tools with which they're most familiar. What they can do is encapsulate all of those file types into a new format called rich music format, or RMF. And RMF will play with guaranteed fidelity on any platform where there's a virtual machine, maybe a year from today.
We're also multithreaded. We can play up to 32 voices simultaneously, and these voices can actually perform separate tasks. So it's important to understand that the audio engine is a software synthesizer that is capable of performing 32 voices simultaneously, but you can make calls at runtime to this engine from a variety of different media. So on the demo that I'm going to show here, which is actually an area of 7Up's web site. [ Shows a slide with a map of the United States, broken into segments each of which features a different set of singers.] I'm going to move my cursor around on this map, and as I move my cursor around, I'm going to be mixing 32 voices in real time.
Now just in case we're having too much fun, it might be rather important, for example, if you're doing this at your cubicle at work, that your company get in contact with you to let you know that you have a visitor at reception. When that happens, the engine is able to manage those different streams in real time. So when I move my cursor over the American flag here, I'm going to simultaneously dim the music and play a voice notification file that came across the Internet. [voice speaks over the music indicating that there is a visitor at reception]
[Applause] Thank you. All of those separate tasks can be managed in real time by the engine. It's important to realize that the calls to the engine are being made by, let's call it the host application. In this case it was the X and Y coordinates of my cursor on an image map. But there are over 50 callable methods in the HeadSpace technology, and just to demonstrate that, I'm going to show you that you can use pretty much anything to call the separate methods of the engine.
In this particular demo, I've got a sample of a soprano singers and over here on the left I have a slider that is going to set the pitch of the sample. But on the right, I have a gif, and by moving my cursor over the gif, I can do exactly the same thing. [soprano voice changes pitch ]
You'll be able to pull in the shock wave movie as an applet, or even a tag of HTML and you can call those methods. It's very important to note that all of these features are going to be available in Java to all of you within the next year.
We're also highly network-aware. And to explain what I mean by that, most of the examples that you've seen so far, most of what you're familiar with on the Internet consists of a single audio file being played from point A to point B. But we can actually create music on the fly by pulling in data from various parts of the Internet.
At HeadSpace, we have some very keen windsurfers, and I'm one of them, and we like to keep an eye on weather conditions in the San Francisco Bay so that when it gets windy so we know when it's time to go out and shred. Well, as you may have noticed, yesterday was incredibly windy, and so we collected some real time data from five weather buoys around the San Francisco Bay and we actually ``Sun-ified'' that data and turned it into music. [ A slide displays a bar chart. Each bar column is a piece of music that represents wind data. They change in pitch and height as the wind data changes .] When things start to get harmonious, you know that it's time to go sailing.
Just to take this a little bit further, I thought that this would appeal to you all. I've put an HTML file here, which I hope you can read on the screen. As many of you that know HTML will see, this HTML file is going to go get a MIDI file from France of a Bach prelude from a server in France, and then it's going to get a shime-daiko instrument sample from a server in Tokyo. If we manage successfully to use the MIDI file in France to play the shime-daiko sample in Tokyo, we're going to go get applause from Redmond, Washington.
I have to say this has never been attempted in real time. This is not a fake. I was a bit worried about Scott's comments about Microsoft's demos crashing. I thought it was bad karma. Here we go. So first, a MIDI file from France. Then we're going to replace the right hand, which was originally written for harpsichord with an exotic oriental instrument, and if we're successful we're going to go get applause from you know who. First the MIDI file. That's the harpsichord. Next we need the exotic percussion. [the sounds merge sucessfully, much applause]
Just to summarize, if you want to check out some of these demos for yourselves, there is a JavaSound kiosk about two floors down from here. I'd just like to say that I arrived in this country on a crest of a wave that was called MTV, and it was a complete sea change for people in my business. I'm getting that old thrill back again, and I'm really looking forward to seeing what you, the developers, are going to do with HeadSpace's technology. But I want to say that it's not just the geeks that get it, it's the artists as well. Thank you very much.
S. McNealy: It's nice to be part of the HeadSpace keynote. [laughter] I'm back. This is.... Miko, maybe you want to just come on up and while we're onto the demos, are you ready to do the next one?
Miko: Yeah, sure.
S. McNealy: Why don't we do the next one? I just thought that this might be a nice introduction. I did a little formula here that I thought was actually a pretty accurate perspective on understanding what ActiveX is: just add a little porting, some memory leaks, and some viruses to Java, and there you have it. Maybe you want to talk a little bit about that.
Miko: Okay, thanks, Scott. Well, the first thing I want to do is boot up this machine. The reason I want to boot it up from nothing is I want to make sure, this is a magician's thing, that there's nothing up my sleeve. This is a totally cold reboot of this system. And we're going to go straight to the Internet. I know that's pretty scary. Certainly scary for me. But there are some interesting aspects of doing a demo like this.
Now this demo requires there to be a floppy disk in the drive. Now one thing that's interesting about this particular demo is that we're running on a lot of different platforms, the last demo was running on the Macintosh. Yesterday we ran on SPARC, and we ran on Netscape Navigator as a platform. The demo I'm about to do for you actually only runs, unfortunately or fortunately, on the Windows platform with Internet Explorer on Win32 in Intel. Okay, so we're going to go to this web page, and we're going to pull up an HTML file. [He types in a URL.]
Hopefully this will load. I'd like to remind you that this is all completely real, completely live, nothing has been cached. We're going to be running something written by Fred McLean, who happens to live near Redmond, Washington. He's the author of something called Internet Exploder.
Now the Internet Exploder actually shuts your machine off abruptly upon visiting a web page. And of course, one fine security feature of ActiveX is that the Internet Exploder only runs on one platform. What's most important about this demo is it's all perfectly true, in the sense that this is all happening, there's nothing cached. This is a digital signature and it's a signed security. So now the components are installing and running.
This is a real web site, but I can guarantee that the moment this demonstration is over, we're going to take the site down, and I'll tell you why. When Fred McLean put Internet Exploder up, he signed it with a Verisign signature. He got a phone call from Verisign, and they said, ``Fred, we're not threatening you, but if you don't take your web site down, you are in violation of an agreement that you signed, and the agreement says that you will not do any denial of service attacks.'' So Fred said, ``Look, this isn't a denial service attack. It's an Internet-based way to turn your computer off.''
This demonstration gives you an idea of the new world of Internet security. If you are open to the Internet, you have the experience of catching all of the innovation all across the Internet. So that means you're capable of catching the brilliance of people like James Gosling, or any of these types of amazing human beings anywhere on the planet. When I was in Bombay, India, I caught up with a guy who had written an entire scheduling software inside of Java, and he had moved it to 1.1 which had only been out for two weeks. This is truly amazing, because these kinds of innovators are the people that you catch when you go into your intranet.
I'll tell you what else you can catch with your intranetthe most virulent hackers out there. For example, Kevin Mitnick. You can catch any of these kinds of people. And this raises the bar of security. It's an incredibly scary world out there in terms of security. You know, in the old days what you essentially had was a situation in which you had to have 80% safety, and you had to protect yourself against things that could accidentally happen.
Well, today you have to protect yourself against things that will purposefully happen. People will be attacking your web page. What's going to happen with the attacks, is that they will be automated attacks, which is even scarier. In fact, I was talking to a security expert who's writing a program called Pandemonium. Pandemonium is a web crawler attack. So what it does is it crawls all over the Internet looking for weaknesses. Think about this for a moment. There is extremely dangerous potential.
Another thing I wanted to talk about is that there are several different hacks that have perpetrated through ActiveX. Do you know about the Moldavia hack? This is my favorite. This is what happens. You visit a web page and it shuts the speaker off your modem, disconnects it, dials a phone number in Moldavia, and then puts it over ISP and charges you. And people have racked up million dollar phone bills over this one service.
You know what? Maybe we should swing it back to Scott. This will eventually finish loading. What do you think?
S. McNealy: Sure. We're getting short on time. Thanks, Miko.
Miko: But this will run.
S. McNealy: This will run. I'm sure it will empty bank accounts.
Miko: It's just a slow download issue.
S. McNealy: It's on a PC, so okay.
What I'd like to do to finish is quickly review what I believe are the compelling features for Java. Obviously we talked about security, and our security model, I think, is one of the real big innovations that we're trying to bring to this world. Microsoft has a different perspective. They think a security model that constrains you in any way, shape or form is not necessarily good. I come from the automobile industry. And we thought that seatbelts were a pretty interesting innovation so when you stepped on the brakes you didn't exit through the window, that staying in the sandbox was okay. And we thought a security model where somebody driving in the car next to you couldn't reach over and step on your brake was actually a pretty interesting environment.
Granted there are some constraints in the sandbox model. But we believe, in the world of network computing, that that's actually pretty valuable. That goes with Write Once, Run Anywhere, the whole conceptthe software reuse, the componentized, no-more-bloatware model is very valuable. We've got ubiquity like we've never hadzero to a hundred million sheets like no other environment has ever reached before.
Somebody ran up to me yesterday and said, ``You never mention it. You never talk about the garbage collection, the `no more memory leaks' feature of Java. That alone is the reason to go to Java.'' And I say, ``It's not fun to stand up and say, ``Hey, we collect garbage.'' You've got a full service API. You've got the network built into the DNA of this architecture. It is a network-centric architecture. A lot of the world has talked about the zero administration client. That's a nice feature. Everybody's made too much noise about the $500 desktop. But the real advantage here is we've provided choice for the world. We've allowed them to choose what microprocessor, what operating system, whatever you need to go get this done.
And the most important feature of all is that you can write your software in Java and it doesn't require one customer to replace their current hardware. This whole concept of ``upgrade, upgrade, upgrade'' is not necessary, because your applications are going to run on all of the legacy platforms. And I don't care if it was a one-day-old legacy machine or a ten-year old legacy machine. We are not asking everybody to get this new feature upgrade. We call it a ``side-grade,'' a side-grade with a Java browser, right to the browser. That's the new strategy, and in fact, with this new product that basically gets rid of all that Microsoft stuff and just runs Java on DOS, we call that a downgrade. You downgrade your computer from this big old Microsoft hairball to a Java client.
It's amazing when you sit down and you go through all of those features, the whole promise of object-oriented programming is only going to happen now because of the virtual machine-interpreted environment. What good did it do to have a reusable component that was compiled to a binary environment in an interoperable networked environment? When you're operating in that virtual safe layer, all of a sudden we are going to see that all of the magic that Steve Jobs has been talking about with object-oriented programming will really take hold. And the new job title, obviously a lot of companies now have a new job title called Web master, but the new job title that's going to become so important in everybody's companies is the ``librarian,'' the Bean librarian is going to be very critical, very important, and play a very profitable, high ROI role in these organizations.
So let me finish now by making a plea for you all to keep doing what you're doing. You can sleep later. Keep writing in Java. Keep doing it. I believe there's a big opportunity for you all here, and I hope there aren't too many MIS types in the room, or maybe too many management types in the room, because what Java is doing with the Web is putting the power into the hands of the author, not the publisher.
In the olden, olden, olden days, think about William Randolph Hearst, a powerful publisher with huge printing presses, huge delivery fleets, who owned papermills, forests, and all the rest of it. He got to a few hundred thousand folks and built this big castle because he made so much money being the powerful publisher. The publisher model has always been incredibly powerful.
The network has changed all that. I mean, Microsoft is the classic example of the modern day eraawesome publisher, lousy author, right? Nobody would ever say that they've written great code to get where they are. They're just an awesome publisher. They've got the brand and all the rest of it.
The Web has changed all that, because you can now write your Java code, publish it on the network for zero cost of goods sold, and zero publishing cost. That is a huge change in publishing of content, and it's a huge opportunity and a huge advantage. Everybody says, ``Oh, you need brand.'' Brand will become more important. Mark Andreesen didn't have a brand when he launched Mosaic in Illinois on a web server, a Sun, I believe. There's no advertising, no promotion. We ran our first TV ads that even mentioned Java last week, to give you an idea. We just put JavaJames Gosling put his stuff on a server, and away you go.
And you have the opportunity to go out and createand it may still need publishers; publishers aren't going to go away! But think about the sports teams, right? In the old days the owners and the coaches made all the money and the players didn't. Now the players are making 10 times the coach salary. Think about that. Think about the opportunity to change that balance of power and maybe even the salaries a little bit. I can just see James Gosling asking to meet me Monday morning.
But I think that's the big opportunity, that's the big chance we all have here. I'd like to finish with two things. One is to thank you all for your support and dedication to the Java environment. We need you. We need the developers. I understand there's somebodyit's like the airplane, it just landed after the safe landing. They say, ``We thank you for flying our airline'' and all the rest of it. I think all of the companies that are promoting at the systems level, the Java environment, want to thank you all, and we know you have another choice and security models. And we appreciate and really do want to support you like crazy. If there's anything we can do, send us email. Get a hold of us, let us know, and keep plugging.
So before I sit down I thought I would do a top ten list. And we put this one together. The top ten indications that Sun has been successful in promoting Java.
Number 1010,000 people turn up to hear you speak and actually stay.
The number 9 reason: Your T-shirts make more money than your applets. Sometimes the truth is very funny.
The number 8 reason: You identify your kids by their class libraries.
The number 7 reason: You're relieved when your Java watch interoperates with Windows and Mac. Well, some of them aren't so funny, huh?
The number 6 reason: You cry when anyone steals your James Gosling action figures.
The number 5 reason: Your boss approves your expenses to JavaOne.
The number 4 reason you know you've been successful: Your porting budget is smaller than your phone bill.
And the number 3 reason: You stop asking, ``Where do you I want to go today'' and you start asking, ``What do I want to get done today?''
Number 2 reason: You, not the chaos club, debits your checking account.
And the number one reason you know you've been successful is that other companies start doing top ten lists at your conference. Anyhow, thank you all.
Miko: So we're here; it took a while to download. I have to apologize for that. But we got here. One reason why it took a long time is because really we had an 8-megabyte CAB file that had to come in from Redmond. So now it's installing the components. Apparently the server over there is a DEC Alpha, so Scott will like that.
We've just downloaded the page and now it's going to run.
Computer: ``There is nothing wrong with your computer. Give it a chance to reboot. We are controlling transmission. If we want to access the hard disk, we can. If we want to change the video, we will. We will control the keyboard. We will control the mouse. For the next few minutes, sit quietly and we will control all that you see and hear. You're about to experience the awe and terror, which reaches from the Internet through the ActiveX
. Please, stand by.''
Miko: So that AVI file was the reason we had to wait. But I think it was worth waiting for. What happens now is a copybut what I'd like to iterate for you is that this is all real. This is all done live. Notice that it can control your keyboard. This ActiveX control can control your keyboard. Now, notice the drive light on the floppy drive. This is real. We just did a slash Q format, which is Quick Format, because the other format took a long time. So we labeled the disk and we're done formatting the disk.
But wait, there's more. I mentioned before that this control actually downloads a copy of WinBatch. WinBatch is powerful enough to control mouse operation, including menu pulls and including any kind of button press. The idea here is it's using the system search facility to search for QDT file types, which is Quicken. And now it's going to bumble around in Quicken and see if it can pull up the user's net worth. So now the hacker, who has seized full control of your computer, is going to pull open the calculator and try to figure out how much money to take from you.
One thing that's interesting is that this is fully scriptable. In fact, as we were doing the walk-in, Fred was doing some interesting little things with the scripting. He was basically changing the script to make it a little more smooth in the flow, and also to calculate how much money to take from you. That's a lot. [referring to screen where totals are being calculated ] Now I haven't actually started doing my taxes yet. But it's getting to be about tax timeso we were thinking maybe we could look up this person's tax records. With just a simple and trivial exercise using this ActiveX control, you can search and open this tax file. So now we're in the program. [the tax file] The scary thing to me is that electronic tax filing is possible, which means that the ActiveX control could trigger an IRS audit, because it could submit totally bogus tax information on your behalf.
What I'd like to do now is call the author of this control onto the stage. His name is Fred McLean. Give him a big hand. [applause ]
And in closing, I'd just like to say I haven't waited that long for a download since the first Java day in New York City when I had to do a live download from Tokyo in a 10 minute talk, and the download took 11 minutes. What a heart attack. I'm glad you got to see this demo because it's a lot of fun. Hope you have fun with Dana Carvey this afternoon. Thanks.
John Gage: All right, we're finishing the keynote. I just want to make a comment about what you just heard. And it's one of those small crusades, not a big crusade, a small crusade, just a piece of email. ISO 7816, the standards for the smartcards. You should know about this. The International Telecommunications Union and ISO still want to charge people money to get standards published across the Internet. So we sent emailI'll get you the email address laterto the Secretary Generals of these two organizations to tell them that the old days are over. We're now a web environment. Standards are free, not 40 Swiss francs, and everything should be published on the Internet.
So that's one small crusade to make all standards published by international bodies instantly accessible to all of us, to all the kids, to everybody, so we know how to make the new world evolve. So that's one step.
The second is, to make this work properly, you must have encryption; we're embroiled in these fights.... The United States is not in a good position on its encryption policy, and so there's a further round to make sure that strong encryption can begin to protect us from many of these things that you've just seen. So we'll talk about that a little bit more tomorrow.
Last, a small homework assignment: A number of people here at the JavaOne Conference do not have a background in networking. They don't really, at their heart, understand how dangerous what Miko just showed us can be. The press often treats this as a bashing contest between Scott and Bill Gates. It's far more serious. This is at the foundation of personal privacy, electronic commerce.
We must do it right as we heard last night. If you're not going to do it right, don't do it at all. Those are the rules for developing Java. Those are the rules as we evolve Java. Those are the rules as we build all the applications that we think transform the world. If you're not going to do it right, don't do it at all. This is serious. This is not a game, and for press people here, ask anyone here who's a programmer how serious it is.
And if you still don't understand it, go to Tsutomu Shimomura's web site, www.takedown.com, and watch the scripts of not-very-bright hackers attacking machines around the world. When Shimomura tracked Kevin Mitnick and captured him (Mitnick's now in prison) the files that were discoveredit didn't make the press very muchthe files that he discovered cached away by Mitnick included SIQ CAD files.
Now does that mean something to a reporter? Not on the surface. What it means is that if I can take a CAD file from Motorola's cellular ASIC that will be manufactured in quantity millions, alter the CAD file, put a hardware Trojan horse in, put them back, and have them compile into hardware, which is then embedded in manufactured devices that go out in quantity millions, I've created a basic insecurity in those devices in which we trust everything.
So we must have a world that allows us to prohibit individuals from moving into files, doing what ActiveX allows you to do. We can't tolerate it. It's not a question of bashing. It's a question of morality and integrity in engineeringbut it's just engineering, it's human beings doing the best they can to build a better world. You're it. Go out and do it. Thanks very much.
Return to top