Related Web Works

----------------------------------------------------------------------------

OMG OpenDoc/DDCF Announcement Transcript

March 21, 1996

STONE: Good afternoon. My name's Chris Stone, and I'm president and CEO of the Object Management Group. Welcome to my new West Coast office. Do you like it? (g)

What I thought I'd do today is gather up some friends and supporters to continue on at least my particular quest, to convince the universe that distributed object technology is where the world's going to go. I've been doing this for now...well, ten years, but six years as the CEO of Object Management Group.

And today is yet another feather in the cap...proves my point. And what I'd like to do is take you through a bit about how OMG views this and views today's adoption of OpenDoc, or what we are going to call the Distributed Document Component Facility -- how that works with OMG's infrastructure, meaning CORBA, object services and things like that.

We'll take you through some presentations from some of the larger supporters and providers of OpenDoc, a demonstration, a Q&A at the end, so please hold your questions until that point. I know you'll be dying to break in, but please hold back. And about fifteen minutes of Q&A and then we'll have a reception.

I'd also like to recognize the fact that on the phone we have some consultants and analysts listening in, so, welcome.

Six years ago when the Object Management Group started, we put together an architecture. And you can see it here on the screen. This hasn't changed at all. This architecture primarily says that the world is going to move towards a component-based infrastructure, right? Towards an infrastructure that says distributed object technology or distributed object messaging will be the platform for which all software will eventually be built.

One could argue that the influx of the recent Internet hype only serves to prove our point: as the pool gets bigger, as more people start to utilize the Internet architecture or ideas, if you will, to start to build software, you're going to have to have a distributed architecture to take advantage of it. Those will be object-based.

We've done what we call the "ORP" -- the Object Request Program --or CORBA, which is formally known. Over a hundred companies in the world today are now adopting that. We've put out fifteen different object services. You can't build distributed architectures without security services. You can't build distributed architectures without transaction services. You can't build them just on a workstation; they have to be platform independent.

In addition to that, OMG's focus is around vertical marketing. People want to build software in specialized areas. Then they want to generalize. They want an architecture and an infrastructure they can build it on. They want to focus on manufacturing execution systems if that's your business. They want to focus on accounting software -- if that's your business.

You don't want to focus on all the stuff underneath, which is what we've basically been putting it through for the past thirty years in software.

Take a piece out of our architecture, we called it the "Common Facility." That's in essence exactly what we're announcing here today. We've adopted as an OMG standard, which we will promulgate throughout the entire world, we will push it into the IETF standardization process, the World Wide Web Consortium, the ISO standards process and the Japanese standards process.

Many of those organizations have already adopted most of OMG's infrastructure specifications. CORBA is an ISO standard, if you didn't know that.

We will then work on other areas in this category. Just so you get an idea, things that deal with financial facilities, repositories. And eventually work towards what we would call common business objects.

Wouldn't it be nice to be able to describe your business and use a set of components and infrastructure software that speaks the same particular language? That's what it's all about.

Now, there's two pieces to this world. The Internet hype has donn an interesting thing to everybody, and it's sort of broken two things apart. First, there's the world of the Internet as it's known today. Many of us who grew up on it know it as something completely different, but as of today this is kind of what it looks like.

Everything at the top of this slide, we're using Web browsers and applications through an HTTP protocol with a variety of different formats that you can utilize, talking to some kind of Web server that uses CGI scripts to eventually get out the stuff that really works, that runs on UNIX machines or IBM mainframes or any mainframe or any environment or NT or whatever.

Then there's what I call the sea of objects -- which is all the development effort happening around distributed object technology. The question on a lot of peoples' minds is, how do these things fit together? How are these things going to fit together?

From our perspective you'll see two enabling pieces on both ends: a CORBA-enabled server and CORBA-enabled workstations. In the server environment, we have a standard called the Internet Interoperability Protocol. It's a distributed object messaging system.

It will, from all the back end systems, tie into all of the quote, unquote "intranet" application development that may be going on in many major corporations. The second is to enable the client -- whether you're using Java as your internal development environment, or Applet, or eventually developing ORBlets. You will use those facilities to eventually develop parts and components. You need an environment to do that in. That's what we've adopted today.

Couple that with a distributed object infrastructure. HTTP is not going to be the protocol of choice two years from now. It has to change dramatically. It has to grow up. It's going to grow up and be a distributed object environment.

When you start marrying these two pieces together and you talk about an Internet Interoperability Protocol around the whole horn you start to see an infrastructure that makes sense. If I can develop a part or a component on any particular platform, I should be able to use a distributed service, a distributed messaging environment, that also makes that real.

That's what it's all about. That's exactly why OMG existed. The strength of the Internet today and the influence of that particular idea and environment only goes to serve the purpose even more.

Now, let's take a look a little bit at what we call the DDCF, or Distributed Document Component Facility, at work. You want to build a component, you want to build a part, in a given particular business environment.

You want to specialize that, whether you're building it using full-motion video or some kind of text environment, or it may be a piculture or some graphic. The issue here is platform independent --not that it just runs on Windows, or it runs on NT, or anything else.

The fact that it runs anywhere, the fact that I can grab parts and components from UNIX file systems, databases, from any particular development environment. They may be OLE based, or ODBC. I want to wrap them into my development environment and base it on a standard infrastructure.

Why? Because I want to guarantee the fact that when I develop a component it's going to work for somebody else. You can't do that today -- anywhere today. Even if you both write it in C++ or SmallTalk or any other language. We can't guarantee that this stuff is ever going to interoperate together.

So your next question would be, what about DCE and what about OLE? How do they fit into this model you've just drawn? Well, to be glib, those are the legacy applications.

[LAUGHTER!

DCE is an example. Most people use it in the sense of distributed file systems or they'll use it...security service. But from a pure document preparation, component building infrastructure? That's not where it's at.

OLE is single workstation, single system. It's going to grow up and be distributed. OMG has just developed a specification that will allow you to take COM OLE on a workstation and completely build a bridge between CORBA and back.

Again, protecting your investment so that you can develop components and applications independent of the platform. Add today's announcement of adopting OpenDoc, and you get a pretty resounding story.

So the message from us is very simple -- before we move on to the next speaker. Component buying today and what people are trying to drive towards today from a Web perspective is nothing more than a distribution channel. There's really not much technology that allows me to build, register, marry these components together, guarantee that they will actually interoperate. That doesn't exist. I can buy this stuff, but I can't do much with it and build a business.

Who certifies what works with what? That's always been a problem in this industry. OMG in cooperation with CI Labs and other organizations to be announced in the future will be building that infrastructure: how you certify these parts to work with each other.

And how many standard interfaces can you name? As far as we're concerned, it's OMG's IDL. From a distributed object perspective -- which is what we firmly believe we're moving towards -- that's what it's going to be based on.

And the bottom line is, if you want to build, you want to register, and you want to certify components that specialize your business, and then give you the ability to generalize them either for in--house use or external sale, it has to be based on some standard platform. And that's what we're announcing today.

Now, what I'd like to do now is show you a video. And if you'll just bear with us, we'll get the video running.

[VIDEO VIGNETTE!

SPEAKER: OpenDoc is the best thing that's happened to software.

SPEAKER: OpenDoc is a cross-platform standard.

SPEAKER: There's no attempt to try to push one platform over another.

SPEAKER: Because OMG is the standards body, you have over six hundred companies that are now saying, this is the way we're going to develop software. This is the way we're going to deploy software in the future, and this is the direction we're going to go in.

SPEAKER: Think of it as a personal toolkit that has the featurest hat I want, when I want them, where I want them.

SPEAKER: The world would be a very boring place if all you had was pepperoni pizza, or even pizza with the works. And that's what you're getting in applications today. With components, you can go ahead and put together the pizza that you really want, the application that meets your need.

SPEAKER: What OpenDoc does is bring us back to that world where the software knows more about us than we have to know about it. SPEAKER: The application of today up to a point are software; but beyond that point they get to be bloatware. It's time to get rid of bloatware and go with component software.

SPEAKER: Building the car, if you don't like your V6, you can take the V6 part out and plug a V8 in. If you want to turbo-charge it, you plug a turbo-charger in and away you go. You don't have to hire a software development or a team of people to go customize that for you. You can do it yourself.

SPEAKER: We have invented a new method of manufacture for quantity production. This system was rapidly accepted...

SPEAKER: [INAUDIBLE! shift. The interest in [INAUDIBLE! has gone up dramatically, exponentially, even. And what we're starting to see is, the number and variety of different pieces of software developed with OpenDoc is going up as well.

SPEAKER: OpenDoc is like Legos -- lets you put pieces of software together as you please.

SPEAKER: What OpenDoc will enable is for small and even mid sized companies to develop expertise in a very particular area.

SPEAKER: It's the beginning of a lot of companies coming together and working on a standard which is really like an intersection where everyone is going to cross.

SPEAKER: A lot of people are trying to deal with the Internet and allk the many different kinds of data types there. OpenDoc makes it very easy to integrate those data types together on the desktop machine.

SPEAKER: The idea is you'll be able to share information with your neighbors. The Internet is another big piece of that. It allows you to share documents all over the place. And of course, if you don't have the same document format as your neighbor, then that's not going to work. OpenDoc solves that problem for you.

SPEAKER: Of course you don't want to have to pull very large programs or libraries over the Internet. You want things to be small. By doing component-oriented software development and software deployment, [INAUDIBLE! just deploy the components that you need [INAUDIBLE! that enables the Internet to be a transport for being able to bring software into your machine, update software that's running on your computer.

SPEAKER: [INAUDIBLE! computers are a universal instrument with which it is possible to revolutionize methods of work in a vast range of fields.

SPEAKER: Yes, but there's no compatibility problems. And that's the promise of OpenDoc. It will work with things that we can't imagine yet.

SPEAKER: The market is just absolutely unlimited and it's going to be revolutionary compared to the way that software's built today.

SPEAKER: This is the Gold Rush. It's a bridge to the future.[END VIDEO VIGNETTE!

STONE: Please join me in welcoming Steve Mills, general manager of IBM Software Solutions.

MILLS: Chris, thank you.

[APPLAUSE!

MILLS: OpenDoc is all about modularity. If you think about the problem that you face in running a business in figuring out how to get a particular job done and get it done quickly, one of the challenges that every company faces is, how do I get my applications built quickly? What's available to me that's going to make it possible for me to get this application crafted rapidly, get it deployed?

And once it's deployed, will it change? Will it adapt to the changing nature of my business?

What OpenDoc is all about is a set of standards that supports modularity. And through that set of standards and the very modular nature of OpenDoc, software pieces produced by one organization, one individual, plugs into the software pieces produced by another.

A very simple idea, but this has not been an idea that's been easy to implement in the past. You see examples of it in the small, around things that are very language specific. Now, for example, I produce a C++ compiler, and I make some pieces you can plug in. And if I can get somebody else to use my specific dialectic version of that language, then perhaps I can get them to produce some pieces that might plug in.

But the possibility for variability and inaccuracy of development and inability to in fact get these things plugged together at time of development, has been a fundamental inhibitor to the expansion of object-oriented technology and the leveraging of the potential for modular component-based software.

What OpenDoc brings to bear is a very stable standard interface, a robust capability for containerizing or fitting component pieces together. It opens up that potential for modular software.

The benefit for customers is the potential to be able to pick and choose the pieces that I need to serve my business. Vendor/suppliers from across the industry can provide components that are going to plug together. I can do it inside of my development shop, and in fact the stability and robustness of the architecture actually literally lets me do it at the desktop by the end user.

That's an extraordinary set of capabilities that's not reflected in any other architecture that's out there today. To the extent that there are other options, they exist largely in the domain of the programmer. They don't extend across to what the end user can do. The robustness and the stability of the architecture is key to delivering that capability.

For the information technology industry, this is a technology that enfranchises developers large and small. Whether you're a small start-up company, an intermediate software development shop, or a large development shop, you can contribute component-ware that can be plugged into other component-ware produced by other suppliers to solve customer problems.

That expands markets. That opens up opportunities. That lets small players come into the market who specialize in specific functions and deliver value.

Now, no discussion of any technology in this industry is not complete without recognizing the myriad of platforms that are out there today. You know, we live in a heterogeneous world, and one of the industry's most popular is Microsoft Windows -- Windows 3.1,Windows 95, Windows NT.

OpenDoc runs on those platforms. In fact, there's a development kit that's out there today. There's some six hundred copies that we've shipped. We expect to ship over a thousand by the time we get into the month of April.

In June we will have our first full-function Beta delivery of OpenDoc for Windows, and we will move towards general availability of OpenDoc for Windows in the fall timeframe. I expect by Comdex that you will see OpenDoc fully generally available on all Windows platform -- a full, complete, fully-documented environment for developers, fully loadable on to a Windows system, be it NT, Windows 95, or Windows 3.1 32-bit.

Now, we have a lot of programs running to help bring OpenDoc out to the marketplace, to make it real for the software vendor community. Within IBM we have a program called Club OpenDoc. Club OpenDoc has some eighteen hundred members already, happens to be one of our most active vendor programs in the IBM company.

We're getting new subscribers every day. We have a Web page set up in support of and promotion of Club OpenDoc. And this is designed to give developers the necessary development kits, documentation and technical assistance out of the IBM laboratories that they require in order to take advantage of this technology.

We're also working on making it possible for small software vendors to get their technology, their capability, promoted in the marketplace, make it available to end users. As we build up a following around the OpenDoc architecture, making components available through catalog shopping and through online facilities such as Cybersource, is a key part of the initiative.

We've had online shopping mall under Cybersource available since last year. Components are register-able and downloadable for a fee on Cybersource. And we expect to expand this throughout 1996.

OpenDoc offers the opportunity for surprise -- surprise in the form of capabilities that as a customer you didn't know existed before. You didn't realize that you could gain access to the components that would allow you to rapidly put together an application and get that application deployed.

Speed to market is a critical driving force in the world today. if you can't adapt yourself quickly to changing market conditions, to global opportunities, to competitors that are out to eat your lunch, you're going to lose.

Information technology is a key weapon for all businesses. Flexible information technology is something that all businesses seek because of these changing external conditions. The ability to build applications quickly, get them deployed, and get them changed based upon changing business needs is the critical issue that corporations face.

Information technology can be a strategic weapon if it's rapidly deployed and adaptable. If it's not, it's a boat anchor on your business. OpenDoc is about making information technology move faster.

Now, what I'd like to do next is to welcome Anthony Brown up, who's going to take advantage of these two gentlemen right up here to step you through a quick demonstration of some of OpenDoc's capabilities. We're going to look at a couple of platforms, and a little bit of distributed computing. And Anthony will explain to you how we put these applications together.

BROWN: Great. Thanks, Steve.

[APPLAUSE!

As Steve mentioned, we live in a heterogeneous world, and OpenDoc is a huge step forward in helping us deliver on the promises that we've just talked about in object technology for a number of years.

And that's the ability to build applications purely from parts, to be able to distribute those across multiple platforms almost in ubiquitous fashion.

What we're going to do today is take you through a very brief demonstration. A couple of things that I'd like to note: one is that we're doing a demonstration that consists of about twenty different parts across four different applications, and then across three different operating system environments.

So we take a very heterogeneous environment -- not only heterogeneous in terms of the number of platform support, but even in the type of components that we're working with. And we'll show how we'll talk about how OpenDoc works with standard OpenDoc parts as well as how they fit into OLE parts or Java Applets, just a number of different component environments that we're having to work with today.

And I'll take you through that very, very briefly. So if I take a look at my first system that I'm working with.... Actually, the first system is an OS/2 application. And what we're going to do here, if I could ask Joe to take a look at the purchase order screen, and look a little bit at how this application is created.

We changed the background of the purchase order screen. What you see is there are actually about seven different components that are built to put this application together. And these components arecoming from a number of different sources. Some of them areprivately or custom-built components, and then some of them like the spreadsheet component are components that are available in the marketplace today.

But purely by embedding these components into an OpenDoc container, we've created a brand new application. In other[INAUDIBLE! we look at the second application -- that being the catalog application. Again, we've taken and embedded information about some wines that we actually have an interest in purchasing.

And we have information about the wine, the price of the wine, and the availability of it. And then if you notice down at the bottom we actually have a Java part that we've also embedded or dropped onto the screen, again from the wine company that we're planning to buy wine from.

So that's kind of where we are in terms of the action where we're starting this demonstration. If I take a look at the second system, it's actually a system running NT. And again, we have another OpenDoc application that is running on NT.

And there are a couple of things I'd like to highlight about that system, the first of which is that again, much like the OS/2environment, we've taken a number of different parts -- again, from different sources -- and dropped and embedded them into this OpenDoc container to create a new application.

A couple of things that you'll also notice is some of the components are basically the same components we saw in the OS/2 environment. One of the things we did, we took a number of the components and in less than a half day, actually ported those components from OS/2 to the NT environment -- again, showing the strength of OpenDoc not only in terms of being able to share components but even in the portability of some of the components.

So again, focusing on the ability to deliver cross-platform applications, and then the ability to distribute those applications. Then if I take a look, again, at my third screen, actually have OpenDoc running on AIX on this particular system.

Again, this is my fourth application, again, a different application, this time running on AIX, but still built using the same concept: the concept of taking different components, different parts, dragging them and dropping them together to build a new application.

So again, focusing on ensuring that we have the cross-platform capabilities of OpenDoc. And now what I'd like to do is take you through a very brief demonstration and focus not only on the ability to build OpenDoc applications using parts doing it in across-platform, multi-platform environment, but also be able to do that' in a distributed environment.

So what I'd like to do is ask Joe to take a look at taking from one OpenDoc component -- this being the catalog -- and actually taking it and dragging it and dropping it into the purchase order component.

Immediately you see that that component has now updated that component into my order for the wine. And as we continue to drop the other components, now we have the option to review the order that we've put together.

And now what I'd like to ask Joe is to send that component, practically place the order. When I place the order, that order is actually being placed to the NT machine -- which is actually the finance manager for this particular application.

So we've taken that order and actually distributed that OpenDoc component and building again on the distributed nature of distributed applications.

So as a third piece of this strategy, what I'd like to do is takes a look at even the NT component and ensure that not only we've built the distribution and the cross-platform, but now we have the option again to take...review that particular order, ensure that it's the way that we want it, ensure that it's accurate, and now we'll actually distribute that again.

When we submit the order this time, it's actually going to do a couple of things. When we submit it not only will we submit it, we will update the original order back to the OS/2 machine and ensure that the order has been processed and sent, but also we're going to take and distribute it to the third machine, which is the good vintage wine company, and update them and process an order for them.

So again, as we take a look at the value and the strength of OpenDoc, we've actually seen a couple of things demonstrated here today-- the first of which is the multi--platform support available for OpenDoc, which we think is very key as we start to build solutions based on the distributed environment.

Secondly, ensuring that we have the support for not only OpenDoc components, but actually multiple components. As we've looked at the industry, we see that there are more and more component models being introduced. And whether it's OLE or Open Doc or Java, there will always be more component types available in the marketplace.

OpenDoc provides an enormous amount of value in allowing not only to work with OpenDoc components but other components as they're being introduced into the marketplace. And thirdly, ensuring that we can build true distributed applications in that environment.

So with that I'll ask Steve to come back up. And that's kind of where we're headed and where we are with OpenDoc.

MILLS: Anthony, thank you.[APPLAUSE!

MILLS: That was, needless to say, a rapid run-through of a lot of capability. And of course, we're going to have some time at the end to answer questions and for people to not only take a look at the demos we have here but also some demos we have running in the other room. So you'll have a chance to take a closer look and ask some questions about these capabilities. Now, with that, to keep our program moving along, I'd like to turn things back over to Chris Stone.

STONE: Thanks, Steve. And to keep it continually rolling, I'd like to.... Please join me in welcoming Ike Nassi, who is the vice president of system technologies and software technologies from Apple.[APPLAUSE!

NASSI: Thank you, Chris. When we started the OpenDoc program, we realized that this was largely a developer play. In order to really get broad adoption of OpenDoc, we had to engage developers very early in the process and we did that.

How does one go about actually doing that? One puts developers into a room...okay? And one feeds them a lot of pizza and a lot of Jolt Cola, and out comes feedback and developer parts. And that's precisely what we did.

Today we have on the order of five hundred thousand OpenDoc CDs in the hands of developers. Five hundred...it's amazing how quickly we get used to ISDN, isn't it? We have developer commitments for five hundred OpenDoc parts to be distributed on the Macintosh platform this year, under Mac OS. It's a major accomplishment for us.

In addition to end users, we've got to develop enough of the component that end users are going to be able to use Open Doc very quickly. And one of the first ways we do that is through something we call "Cyberdog." Cyberdog is an Internet-based user environment that brings the traditional Apple ease of use out on to the Internet.

And there's several different components, and we'll be showing you a demo of that very quickly. It's our intention to very broadly distribute OpenDoc. OpenDoc is available on our Web site today. And you can download it, you can try it out. And it will be shipping later this year.

One of the advantages of OpenDoc is that it allows us to bring technology to market much more quickly than traditional monolithic applications. And what do I mean by that?

The problem with traditional monolithic applications is that Apple or another company might produce some interesting technology, and that interesting technology then gets shipped off to a developer. And the developer goes through a long development cycle and puts that technology into their application.

And by the time you turn around, there's a fairly long period of time from the time you originally developed the base technology to the time the technology gets into the end user's hands. One of the very interesting things we can do with Open Doc and that

OpenDoc enables is that we can shorten that cycle tremendously. We can in fact provide OpenDoc parts for key pieces of technology so that the user does the integration as opposed to a third-party developer having to go in.

So the key technology pieces that Apple can provide are things like QuickDraw 3-D, QuickTime, QuickTime VR, QuickTime Conferencing, speech, and all those things can be packaged by Apple as parts and then integrated in a variety of ways by developers or by end users into a particular solution.

But even that is really not enough. One of the things that I'm pleased to tell you about is that ClarisWorks will be an OpenDoc container sometime later this year. And once OpenDoc is a container, that provides a very rich environment with a common set of parts in which people can add additional flexibility.

ClarisWorks, for those of you who don't know, contains things like spreadsheet, word processor, drawing package, and so forth. And so all of the basic fundamental pieces will be there for an end user application, and it's a good opportunity for a user or a developer to add parts to that application and really create a custom application.

If you look at the OpenDoc program in three phases, phase one was developing the basic platform, which we shipped in November. Phaset wo enhances that platform, works with developers, and gives us an opportunity to develop some end user parts -- which we're doing.

But, if I look a little further out, you might ask, well, what's going to be the next phase of this process? We finished phase one, we're in phase two. Phase three, you're going to see a much closer integration of OpenDoc with the Macintosh operating system.

And there's several things that we're doing, several things underway. One of the things is that we'll be integrating our mail strategy in with OpenDoc and providing that. I don't think we'll be demonstrating that in a lot of detail today, but I think you'll see that there's some very exciting things Shawn is going to show you right now with a Cyberdog demo. And with that, I'll turn it over to Shawn Kruden. And let's see the Cyberdog demo. Thank you.

KRUDEN: Thanks, Mike. My name's Shawn Kruden, I'm a Cyberdog evangelist at Apple, and I'd like to take a few minutes to show you the first end user solution we're going to ship on top of OpenDoc. It's code name is Cyberdog.

It's a collection of OpenDoc parts to do Internet navigation and browsing. It wasn't set out to be an OpenDoc solution. They designed it with the goal in mind of trying to bring the ease of use of the Macintosh to the Internet experience -- because right now you need to be a little bit savvy about what's going on in the Internet to take full advantage of all the information that's out there.

Their design goal was ease of use and seamless integration in the user experience in navigating the Internet. They very soon discovered after they started designing the product that the Internet moved so quickly they'd need to be able to really easily replace functionality over time.

So they wanted to use a component software architecture to implement it, and of course they chose OpenDoc. So what I'd like to show you is that seamless integration experience, and also then show you some of the benefits they get just because of the fact that they're OpenDoc parts and how that functionality can then be embedded in other more traditional applications where we haven't seen Internet content before.

One of the things that we're doing in an attempt to help users get on the Internet is to organize the information and give them a starting point. So there will be some sort of starting document. This is sort of a Beta-level quality, so it will probably change before it ships. But to give you an idea, we're trying to give people easy access to discussion groups, e-mail, to a Web browser that has a number of hot spots in it and what not.

There's also two facilities that override all of the services within Cyberdog, and that's the log and the notebook. I'll show you the notebook in just a minute, but let's take a look at the log first of all.

The log is a history mechanism within Cyberdog. Every Cyberdog registers with it as soon as you visit a new site, and it keeps track of everywhere you've been. So in alphabetical order, if we were looking for something about OMG, we should go down in alphabetical order.

You can also look at it in chronological order in which you visited the site. Or, you could look at it in hierarchical order which shows you the way in which you surfed to that particular place on the Internet.

So you can think of it sort of as breadcrumbs through the Internet. Anything that's in that log can be double clicked on to go immediately back to it. And Cyberdog's very clever about knowing what type of service it is, automatically loading up the browser that belongs to that particular data type, passing the URL in, and then taking you to that particular place.

So if we go back to the OMG pages, I can show you how we're going to do integration with the Finder, the desktop environment on the mac OS. I can take any of these icons whenever I see them in the log or the notebook or within a browser window itself, and drag that to the finder. And then at any point in the future I can double click on those and it will do the right thing, which is basically take me directly back to that particular spot on the Internet.

So if I come out to the Finder, I might come back to my machine the next day and I want to go directly back to that place on the Internet, I simply double click on this similar to an alias in the Mac OS today, which knows how to get to that particular place.

Cyberdog's very smart about knowing about what different types of URLs or locations things point to. This happens to know that it's a Web site, so it finds the preferred Web browser from within Cyberdog, loads that into memory, passes the URL to it, and takes you there directly.

So as an end user, I don't have to be so savvy as to know which browser I have to launch and then type in the URL to take me there. I can just be given this Cyberitem -- is what we call them -- by someone else and just go to a location without having to know a lot about exactly what's involved. So here I've got something that takes me to the OMG home page and talks about membership information.

So let's pop out of here, and I want to show you a custom document that I've built -- because I think the most exciting part about Cyberdog isn't necessarily the seamless integrated product that we're going to ship to allow users to browse the network, but it's the benefit that the browsers get because they're OpenDoc parts.

So let me get this launched. What I've created is a simple document that has some text and graphics in it. So in the future, it could be something like ClarisWorks, for example, or any application that becomes OpenDoc enabled. And I've created a document that the OMG might use when they're trying to recruit new members, for example.

So I've got a little bit of text describing what the OMG is all about and why you might want to be a member. And embedded inside of this document just because of the fact that it's an OpenDoc container, I've got a tool that's provided with Cyberdog called a Cyberbutton up here.

Cyberbuttons are really simple. You can put a graphic on them and then wire them up to someplace on the Internet that Cyberdog knows how to get to. So I also should mention at this point in time, you can see down here I'm actually live on the Internet. I've got a modem, 28.8 connection to a Netcom local phone number -- so hopefully the line will stay with us for the rest of the demo] But, I'm live on the Internet here.

The whole point of this document is I can give this to end users who might not be that savvy about the Internet and provide access to information that's of value to them. Instead of sending them a flyer with a URL that tells them where to go in their browser to actually get someplace, I can just send them a word processing document or an e-mail message that has this information embedded right inside of it. So it's sort of one-stop shopping. They don't have to be as savvy as you have to today in order to be able to navigate to the information and take advantage of it.

So I've got this particular button wired up to another Web site, which is the home page for OMG. So I'm one click away from here. The next thing I want to do is actually show you how you can embed this and how easy it is for users to create these documents.

This isn't something that a programmer needs to be able to do; Cyberbuttons are very easy to create. And the next thing I want to show you is how you can actually take these browsers and actually embed them inside of your document.

So this is an OpenDoc enabled application. Any of the browsers that I've been showing you are in themselves Open Doc parts, and therefore can be embedded inside of any OpenDoc enabled application.

So I'm going to open up one of the other facilities that comes with Cyberdog called the Notebook. The Notebook is the facility that allows you to keep track of your favorite places. And anything that Cyberdog knows about can be registered and saved off in your notebook.

So, like in the log where I was showing you have those aliases that you can double click on to take you directly to any place on the Internet, anything from the log can be dragged into the Notebook. Or when you're in a browser window, there's an icon that you can drag out of that window as well to the Finder in the Notebook.

So I've got here a collection of sites on the Internet that would be of interest to someone who's going to be joining OMG, for example. I can double click on anything right from here to go directly there, but what's more interesting is if I drag and drop this newsgroup, for example, inside of our document.

What this is actually doing is going to our news browser and embedding the browser in place inside of my document. So when I get this document as an end user, I don't even have to know that it's an Internet newsgroup. I just see that there's a list of discussions and documents about a particular subject that I can directly manipulate to get information and contribute to a newsgroup.

So for an end user who hasn't done a lot with the Internet yet, someone can take the time to research areas of the Internet that are of interest to a particular set of users and create custom documents for them where they have one-stop shopping, things that are very valuable to that particular group of people without them having to know exactly what it is that they're using necessarily.

So here I now have a document that I could save off in this state. The next time it's opened, this news browser will reestablish its connection to that particular place on the Internet and then refill that pane with all the latest discussions that are going on in this particular group.

So if I scroll down here you can see that we have some threads going on a number of different topics. Java seems to be hot these days. If I double click on any one of those, Cyberdog again takes over, it knows that it's a newsgroup, it will automatically load up our news reader and then show me that particular discussion area.

Our news reader and our e-mail system are integrated, so it opens up...it's the same part that you use for doing both -- which is nice because as soon as you want to reply to a particular discussion point, you're already in the e-mail browser and it automatically addresses everything for you and what not.

I wanted to show you how to build one of those buttons because it's very straightforward. In OpenDoc we have a concept called "stationery" which is basically a template of a part, so give me anew one, I can just drag some "stationery" in here to get a brand-new button.

We're integrated with the System 75 Scrapbook, so I can just drag and drop a graphic on top of it. And then I can get it wired up any number of ways. Any time I see one of these icons -- we call the Cyberitems -- it can be dragged and dropped on to a button wired up to where I'd like to go.

So I can drag and drop out of my notebook, or my log, or any of the browser windows that I might have up. And it's as simple as that. So now I've got a button that's embedded in any OpenDoc enabled application that can be wired to any particular place on the Internet.

So I can very easily now create custom front ends to the Internet to provide people access to the valuable information that's on The Internet, and do it in an application that they're very familiar with. So we can take information that we can find and put it together in something like a word processor or a spreadsheet for example, anything that's OpenDoc enabled.

So, I'd like to show you just one more thing, and that's our integrated e-mail system. I've got a couple of contact names from OMG. Here's Chris Stone's electronic mail address. And when I double click on any of these things, it will open up our e-mail browser which we saw just a moment ago in the news area, automatically address it to whoever that's going to, and then I can start typing it in.

One of the things that we've done which I think is very nice is we're supporting MIME, which is an Internet standard for mail. so I can say, Just saw your press release, liked the type, looks great.

One of the things with MIME support is the ability to actually support styled text. So unlike most e-mail packages where you're sort of restricted to Courier 9 point or Geneva 10 point or something like that, we can support any font that you'd like, such like a word processor.

So I can make it bigger and have styled text, for example. MIME also allows you to support the integration of graphics. I can take this same logo and drop that in as well. And because these Cyberitems are just any other kind of object, I can also embed those.

So I have a Cyberitem I've saved in my favorite places about OMG press releases, and I can drop it right into my e-mail message and send it off.

If this goes to a system that does not support MIME, it will strip out all the things that don't make sense for it like the style text. It will take the things that have been enclosed in the middle of the message and make them attachments instead so that people can still get access to it. They won't have the full-featured experience that we would have on a Cyberdog system.

If somebody has Cyberdog at the other end and they see this, they would get this e-mail message and then be able to manipulate this directly, pull this out and put it in their notebook, or double click on it directly within the e-mail message and it does the right thing, knows it's the Web site, pulls up the Web browser, and takes me there directly.

So we see some really nice integration between all the different pieces in the Cyberdog product itself. The way it's different, I think, from what we're seeing in traditional browsers is a couple of things.

One is that it's much more seamlessly integrated. But our philosophy about the Internet is different from the traditional browser community in that browsers seem to be going down the path of taking every kind of piece of data and putting it inside the browser. And we're taking completely the other aspect of it, which is taking everything out of the browser and putting it other places.

So anything that is a Cyberdog part and supports Internet information can be embedded in any OpenDoc enabled application. So we're looking at supporting things like Netscape plug-ins and Java applets within the Cyberdog environment, and then that would allow you to take those things and embed them in different places, not just your browser. So you can imagine running a Java applet from within a ClarisWorks document, for example, in the future.

So, that's what I wanted to show. Hopefully you'll get a good feel for...look at Cyberdog as a case study of Open Doc, what can you do with OpenDoc and what kind of benefits do you get out at the other end.

We're really excited about it. It's on our Web site right now, www.cyberdog.apple.com, at a V2 level. And it will be going golden master at the end of April. We'll be distributing that at our Worldwide Developer Conference in an SPK format.

And then for our end users we'll be shipping a product at Augus tMac World that will be made available freely to everybody on the Web, and we're really excited about that. I'd like to turn it back to Mike to close it off.

[APPLAUSE!

NASSI: Well, I think you can begin to see Cyberdog as an example of an OpenDoc application. With that I'll turn it back over to Chris.

STONE: Thanks, Shawn. I don't remember seeing you breathe throughout that entire demonstration] Our next speaker is Bruce Cleveland who's the president of Component Integration Labs. Bruce?

[APPLAUSE!

CLEVELAND: As I think somebody famous once said, It looks like really insanely great technology. A couple of interesting things: first and foremost, I believe that this is actually going to be a very significant event in the software industry, and that is by the OMG adopting the OpenDoc specifications, what we've really done is essentially taken the traditional server-based CORBA developers and united them with what have traditionally been the high-volume client base distributed or component software developers.

And in essence, this unique combination of client and server is going to form the basis for a brand-new distributed component software platform that we believe will essentially ratify or essentially change the entire way in which we design, develop and deliver software in the future.

I think this is an important point, because I fundamentally believe that until we have a ubiquitous new platform that is agnostic across operating systems and environments, it's going to be very difficult in order to really achieve the design objectives that the OMG has set forth.

So that's what this mission of CI Labs is all about, and that is to adopt, license, register and validate distributed component technologies that once integrated become part of a distributed component software architecture.

CI Labs has formed the basis of a distributed component software architecture that we also call DCSA. And we work synergistically in the formation of that by adhering to or identifying the object standards that the OMG specifies, and then looking into the industry and finding technology that adhere or conform to those APIs.

The result of which is a very synergistic relationship where we can have technology made available for those people who desire to build or implement a platform, and those component developers who are looking for a ubiquitous platform that they can be ensured that all services are available on.

So as it refers to this distributed component software architecture, OpenDoc is the cornerstone technology to do that in that it provides the compound document and components services part of the architecture. And there are other technologies including IBM's DSOM which is a CORBA-compliant distributed ORP that really round out the DCSA environment.

In total, there's been over a hundred million dollars' worth of investment made in these DCSA-based technologies. And CI Labs has been authorized to license those technologies freely to any member of CI Labs. And it's a very, very important note.

So in order to get these to be adopted widely, we have to make sure that interoperability is the linchpin. And to do that we have to make sure there's a set of validation test suites at the component and at the platform level.

One of the things that we're committed to doing by this summer is delivering a component software validation product both in Mac OS -- and on OS/2. As soon as the other DCSA technologies are made available on other platforms, we will also provide the component software validation product for those as well.

The result of all this is that we're seeing a tremendous growth rate in terms of our membership. In fact, over the last six months membership has increased ten-fold -- from thirty to over three hundred -- and rapidly growing each and every day.

We also want to make an important announcement, and that is we've added a brand-new sponsor -- Just Systems -- to our fold. And I believe that this is a critical announcement because they happen to be the world's largest Japanese software vendor. And while you may not know their name here, in Japan virtually anybody who owns a computer system uses a Just Systems product.

They create business, connectivity and groupware type products but they're also the leading provider of an input method called A-Talk that nearly every system vendor OEMs.

So we believe that having Just Systems as part of our organization is a critical step forward as we move outside the boundaries of just the US. In addition, because of all this activity a lot of task forces are forming along with Just Systems. We want to invite every other developer into the CI Labs organization.

So we have a special offer that we're making to all existing OMG members as well as anybody else who wishes to become part of the OMG on or before June 30th: and that is if you do, we're willing to give you a complimentary membership inside of CI Labs at the subscriber level.

And in addition to that, give you what we think is a very important document in the form of a book which describes a lot of the concepts, the tenets and the issues that we've been describing today: how OpenDoc works with CORBA, how they extend and complement each other, how all this is going to work with OLE an other software component architectures.

And if you wait around any longer, we'll probably have a set of Ginsu Knives for you as well, but...] I would like to take the next two minutes to recognize somebody who I think has contributed a substantial amount to the software industry and the object technology world specifically.

Sometimes we don't recognize the kind of stamina that it takes, the kind of intestinal fortitude that's required to sort of go against the grain when you have a vision about where things should go, and standing up to people who say, "it can never be done," "it will never happen," "no one will embrace this kind of a technology direction."

And I don't think that we do enough for the people who are willing to take the risks to make that happen. And what I'd like to do is recognize Jed Harris. Jed Harris is a distinguished engineer and scientist at Apple. He's also the original founder and former president of CI Labs.

And Jed has done a tremendous amount to move the technology and the vision forward. This relationship with the object management group, our relationship with Just Systems, are primarily gaited on and were depended upon Jed's vision and ability to be able to describe technologies that no one had really heard about or used before in a way that was meaningful. So with that I'd like to invite Jed up here and present him with this award.[APPLAUSE!

HARRIS: Thanks very much. Well, this is a great step. I think this is a real fulfillment of the vision we had in defining OpenDoc and in defining the CI Labs business model to make this an industry standard that really works well for the developers and for the end users. Our goal was to bring those together and make something that would work well across all the range of function that people have been talking about today.

And making something that are industry-standard explicitly through the OMG is a major step in that direction. We're very proud of that. Thanks.

[APPLAUSE!

CLEVELAND: Thank you, Jed. And with that we invite everybody to join CI Labs and help us shape the future of software.

[VIDEO VIGNETTE!

SPEAKER: No matter who you are, if you're a student, if you're at home,, if you're in the workplace, OpenDoc responds to the flow of your work -- how you move from thing to thing in order to get something done.

SPEAKER: It's getting too complicated the way it is now. We wanted [INAUDIBLE! component. And OpenDoc really is the only framework that's out there that allows us to do that.

SPEAKER: It would be possible to integrate systems that were based on entirely different desktop computers using entirely different networks and have them all talking to the same back end using OpenDoc.

SPEAKER: OpenDoc is really the glue that pulls this all together.

SPEAKER: We're on the verge of a whole new level of software development, a whole new level of software choice, for users.

SPEAKER: OpenDoc is the only technology component around -- the only one that gives you the cross-platform capability, the openness, and the way to build components that are going to actually survive in the marketplace.

[END VIDEO VIGNETTE!

CLEVELAND: I have a very simple job and I've been telling you this when I started out the presentation. And that simple job is to fly around in airplanes and talk to as many people as I possibly can and try to convince them that distributed object computing onrushing standard infrastructure in distributed objects is the only way we're ever going to get to a component software industry.

I've been doing that for seven years and I'll keep doing it until it gets through everybody's thick head that that's the only way to do this. And what I'd like to end up this particular presentation on is, you will see in the future more and more vendors, suppliers, component builders, individuals, whatever it takes, building more and more specialized applications using the products you saw demonstrated here today, using products that you didn't see demonstrated today -- all using the infrastructure that OMG has put in place.

And I want to thank you for coming. Now, here's the fun part --which is the questions. And hopefully you'll get some answers to your questions. So I'd like to open it up to the floor. There must be some questions from this audience. You are not a shy bunch, I don't think.

If you have a question, by the way, direct it to a specific speaker. Please do so. Question?

QUESTION: I'm Dana Gardner with InfoWorld Electric. You've talked about shopping for components and shown how components go in and out of applications. How does the business model follow the technology in that how do you convince the vendors that the components that they've made and then used, customized or elsewhere, that they'll get paid for that? And how do you convince the user that what they're buying is only what they're paying for?

[LAUGHTER!

CLEVELAND: You want to try that one more time, please?

QUESTION: I'm sorry. If you're a developer and you've got these components and they're being picked up and used like the pizza model -- different pieces -- how do you get paid for just your piece? And if you're the pie builder, how do you only pay for the pieces that you're putting on it?

CLEVELAND: Okay, that's a two-part question. Let me see if I can put words in your mouth. The first part of your question is, if this all works and it's all well and wonderful and we're building components that are based on an infrastructure and industry standards, how do you make any money at this. Right? How do I build this stuff and make any money at it? That was the first one.

The second one was, given the fact that I can make money at this, how do I ensure that people are actually going to buy it?

QUESTION: Using these components, how do I satisfy my shareholders and my CEO that I'm only paying for the components that I'm using rather than buying the entire application and only using ten percent of it, which I might be doing now?

CLEVELAND: Let me take a shot at the first part. First of all, there's not much of a different business model from putting all this stuff in a great big huge shrink-wrapped box and charging five, six, seven hundred bucks for it versus taking one of those as a part or a small component and putting it on either the Internet, a Web page, or whatever it may be in some electronic form and charging some number less amount of money for it.

The business model still applies. The problem is, in an electronic form, almost anybody can grab it. And you're talking about then the reuse of that over and over again down the line, right? Let's say thirty people begin to use that.

But it also interjects a couple of other issues, like quality of the software. It interjects things like bug fixes -- who broke it? Particularly when I've got twenty or thirty people down the line that have been using this thing. Whose fault was it when it eventually got to me?

As far as I can tell there's no clear answer to that particular question yet because it hasn't started. There's a lot of distribution of software components in electronic form, and they all seem to fit and they're continuously following the same business model of price it for what it's worth in that particular piece of software.

I think over time what you're going to end up seeing is as these parts become standardized, as they start to use interfaces that actually can talk to each other, you will start to seek their market value for certain categories.

There will be pieces of software that fit into different categories. And those categories will have a certain amount of value attached to them. If you look up in a auto magazine, you see tires. They all have sort of some...they're all hopefully the same price. They all do different things, and some of them are very specialized.

The same thing I think is going to happen in the component marketplace. Steve was talking about Cybersource as an example of a mall, if you will, or a shopping environment where that's actually going to take place: software communications. And expect that to happen on a standardized platform over the next couple of years.

QUESTION: So does "standardized" translate to "commoditized"?

CLEVELAND: Absolutely. "Standardized" generally relates directly with "commoditized," which helps to drive further development, which helps to drive further specialization in given areas and given marketplaces.

If you believe the world goes in this particular direction, then you end up with a verticalization which will start actually building software that's very verticalized, very specialized marketplaces.

Now, OMG's organized around telecom, [INAUDIBLE!. Most of the vendors in this room are organizing their businesses around very specialized markets. That's where it's all going to head.

UNIDENTIFIED: Let me just add one thing to that. OpenDoc...it actually feels funny given that Jed is sitting here and he invented a lot of this stuff. But, OpenDoc is built on an object-oriented foundation, and a lot of the parts of Open Doc are capable of being overridden and extended in a lot of ways.

It may be the case that the traditional methods of distributing parts are not sufficient for the particular environment. We don't really know, we're going to get there, and we'll find out.

But if in fact that's the case, if in fact the mechanism that we use today -- the traditional licensing mechanisms, the distribution mechanisms we use today -- are not sufficient for this, it's very possible to override parts of OpenDoc and put in things like metering code, without changing a lot of other things about OpenDoc. It's possible to do that. An opportunity]

QUESTION: Michael Parsons from the IBG News Service. Novell's been talking about their Netware Directory Service as being a way to map and control all these different software elements flying around, distributed software architectures. Do you think that's credible?

CLEVELAND: Do I think Novell's NDS is credible? Is that what your question is?

QUESTION: As a container and as a way of controlling, organizing, licensing and distributing...

CLEVELAND: Oh, as that particular model. Do I think it's credible? I don't know.

Time will tell. It depends on whether you buy it or not in theuser community. So I think it's credible to say that NDS and the services that Novell is going to implement over time comply with, let's say, a standard infrastructure like an IDL interface. I think that's going to help Novell move towards an area where you accept their services more. In other words, you find theirservices more palatable on different environments, not just an Netware MVS environment.

So I think what's going to happen with Novell and other companies is they're going to follow this same kind of model where they'll be developing services that they can sell as almost a component or apart -- a component service -- that is available for use in virtually any environment. Same model.

QUESTION: A lot of you would probably say it's a weighted question, but to a lot of people that we see here that we talked about compatibility with Windows and OLE objects, and since the basic structure of OpenDoc itself and the power of the container of OpenDoc actually has to be dumbed down to work with OLE and the way that it works.

[LAUGHTER!

Are we going to just limit OpenDoc's growth and the ability ofOpenDoc by keeping compatibility with OLE and Windows?

CLEVELAND: I'm not going to answer that question. I thinkthat's probably better served to have IBM or Apple try to attempt to answer that.

[LAUGHTER!

UNIDENTIFIED: I'm trying to stay out of trouble.

MILLS: Let me take that. From my perspective the way to deal in today's market given the number of participating vendors, customers, users, the myriad of technology options, is to deliver technologies that enfranchise users but come at the market from different directions.

And there's certainly a direction in the market driven by Microsoft, that has captured a lot of attention of Independent Software Vendors and developers. The interoperability that's provided, the ability to wrapper and contain OLE components, I think is an important leverage point for the OpenDoc architecture.

You know, it's not our intention to turn this into a set of binary decisions for the customer. The fact of the matter is as a customer you're trying to run a business and you want to take advantage of whatever pieces of technology are available to you, and you'd like to get the pieces to work together.

We're focused on doing that. Microsoft's been very much focused on satisfying needs on their own platforms and not elsewhere, and some of us are in a cross-platform business and therefore want to enfranchise a larger constituency.

So, you know, there are always trade-offs that you make when you try to go for higher levels of interoperability. Certainly native OpenDoc components have a level of functionality that go beyond what we see in today's OLE architecture.

So emphasizing native OpenDoc is certainly something that we're doing through the Club OpenDoc and other vendor programs such as Apple's program.

QUESTION: But that would lead to a follow up question, if I may, being an obstinate person. Instead of forcing OLE and all them to come up to OpenDoc standards, then when OCX comes out, which is an interim step between OLE and OpenDoc, we will then support that instead of setting a guideline and basically a standard, because OMG is a standards group, have...force Microsoft to bring their product which it was capable of bringing it up to that level?

MILLS: Yes, well, we're sort of watching the evolution of this technology, right, from DDE to OLE 1 to OLE 2, the automation level, OCXs, and most recently, their Active X announcement. So you really have to keep close tabs here on the ever-changing names and...

QUESTION: Will you still call it CORBA?

MILLS: We still call it CORBA. Haven't changed the name.

QUESTION: Thank you.

QUESTION: Mike Kennedy, META Group. I'm looking for a little bit of help in trying to position it with regard to the IOP. As a developer, when will I use the IOP versus OpenDoc? And also in terms of positioning it versus complex multimedia management that's managed by object databases? How do they all fit together?

CLEVELAND: All right, let me answer your first one. The IOP, in essence, is the underlying protocol that you would use to distribute for the remote method indication, if you would, or the distributed message. That's in essence what the IOP does.

OpenDoc, doesn't matter where it is -- where the part or component is. It could be, most likely, on your workstation but there's no reason eventually that couldn't reside in any particular server environment or HTML-like environment, if you will.

Think of IOP as nothing more than the messaging mechanism. That's exactly what it does. But the most important part about that is it guarantees the interoperability of the part. It guarantees that when the message from one end gets to the other point, and they happen to be designed by two completely different vendors or individuals, that they will interoperate with each other. It goes right down into that low level protocol.

And that's exactly what IOP was designed for. You will see thevendors in this room supporting IOP. You'll see lots of other vendors as they come out with their ORPs, give you the capability of using OpenDoc or the DBCF facility. That's exactly what it's all about. So you are completely free from the underlying complexity.

Now, the second part was...?

QUESTION: [INAUDIBLE!.

CLEVELAND: Complex data structures managed by object databases-- again, IOP acts as a messaging mechanism, whether you're talking about an interface such as a SQL interface or as SQL grows up to be SQL 3, et cetera, or you may be talking about projects from one of the larger database vendors like...let's pick one, Fedona, from Oracle, as an example...

Or some of the projects from some of the object database companies. They still need an underlying messaging infrastructure. You will start to see IOP being used as that messaging infrastructure from those companies, whether they be large relational databases or more complex data structures from object databases.

In addition, the object database companies will change dramatically to be more application development companies -- not necessarily object database. That's a technology, not a market.

QUESTION: Yes, Steve, a question for you. Norman Leech from PCWeek. Actually, two questions. The OpenDoc for Windows coming out in the fall is, I take it, 32-bit Windows?

MILLS: Yes, we're delivering a 32-bit.

QUESTION: Any plans for 16-bit?

MILLS: No.

QUESTION: It looks like VisualAge is going to be the only tool that initially supports it. Are you talking to any other tool developers to provide frameworks or anything to build up those parts?

MILLS: Apple has tooling today to support...

QUESTION: For Windows.

NASSI: We've got what we call the OpenDoc Development Framework which actually is an object-oriented cross-platform development environment that will allow you to build parts both for Windows and for Macintosh.

MILLS: The IBM Open Class Library including the compound document framework is openly licensed to any other vendors that want to license it, and it provides support for both OLE and Open Doc through the same compound document framework. We have a variety of different discussions underway within the industry for licensing that structure.

QUESTION: Any companies you can name?

MILLS: Not yet. Actually, it was just delivered -- as I think you noted -- here within the last couple of weeks.

QUESTION: Jeff Bartlett with IT Solutions, a question for IBM: why did you decide to drive the OpenDoc compatibility for Windows, and how did you do that without going through Microsoft?

MILLS: Well, of course, compatibility is an interesting challenge. The fact of the matter is that there's enough of the externals of OLE that are evident in the documented syntax -- the eight hundred some odd expressions that make up OLE, that we're able to look at what those functions are calling and map them across into the OpenDoc structure.

You know, that's a very do-able implementation. As always, one has to keep a close eye on whether or not they change the underlying capabilities of those interfaces. In other words, if they start to change the lower level service calls that are associated with that function, then we'll have to change and adapt for that through time, to maintain interoperability.

QUESTION: Are you going to be sort of chasing Microsoft for the foreseeable future to keep OpenDoc compatible?

MILLS: No. As is the case with any.... OLE is a language, right? It's a language that's unique to run times that they deliver. Its externals are documented. Obviously to the extent that OLE grows in popularity, the stability of those externals must be maintained or your applications will no longer run. You'll go through iterative constant rewrites -- you as a customer or you as an Independent Software Vendor partner of Microsoft.

So the very forces that drive Microsoft in terms of trying to make OLE popular across the industry in turn cause them to have to standardize and stabilize its external interfaces in order to promote the vendor and customer base forward.

So I'm not particularly...I don't lose any sleep over it. I am not trying to build a binary compatible execution environment inside of a [INAUDIBLE! container for OLE. I'm merely trying to map its interfaces so that I can intercept the functions that it is calling and deliver interoperability in the Windows environment for OLE users and OCX components.

But that's very different than trying to deliver the equivalent of OLE. We're delivering something that captures what OLE does, and then goes beyond what OLE does. Those are very different solutions.

UNIDENTIFIED: Also from OMG's perspective, Microsoft doesn't expose the description language, which is exactly what we do. There on we expose the description language is so that you can standardize on that API.

UNIDENTIFIED: That's in essence what you map into, which is kind of what Steve was talking about. Whereas from a Microsoft perspective, they build products, right? Period. You don't see much of an interface, you don't see an exposure of a particular API. And we're willing to deal with those kinds of issues.

That's partly what we...our job, and what we've been doing with Microsoft over the past umpteen years but have recently got them to agree to work with us on this COM to CORBA thing you may have heard about. In essence, it will be somewhat our job to make sure that the products that are developed that use some form of an OLE integration can actually speak OMG's IDL and give you some sort of interoperability. We will maintain that specification that does that work.

QUESTION: Henry Bortman, Mac User Magazine , yet another question about OLE. I was under the impression that when Novell exited the scene that there were delays in the OLE interoperability, and everyone's sort of talking here as though it's the. Is it in fact in OpenDoc 1.0 from Apple and will it be in what IBM delivers later this year? If not, when?

MILLS: All of the function that was worked on by Novell has been passed across to IBM. So everything that they built, we have. And we're now extending that to the final deliverable for OpenDoc. Sonothing's been lost.

UNIDENTIFIED: One of the things that we're looking to do is get some feedback from developers on exactly this subject. And depending on that feedback, we'll tune the development schedule to that feedback.

OLE interoperability is a very important part of the product. it will be there, it's not there right now.

QUESTION: You don't have delivery dates for it yet?

UNIDENTIFIED: I'm really shooting for late summer.

QUESTION: How will OpenDoc generate opportunities or new markets for hardware vendors, systems vendors?

UNIDENTIFIED: Her question was, how will OpenDoc generate or help create opportunities to generate revenue for hardware system vendors. Well, you've got a couple of them in the room here, I'll let them answer.

UNIDENTIFIED: As I said earlier, I think one of the things that Apple is known for is introducing new, innovative technologies. And OpenDoc is a very good vehicle for trying to accelerate the process of getting those technologies into the hands of our customers. So by packaging some of these new technologies as OpenDoc parts, we accelerate that process.

MILLS: Besides the notion that more software drives more hardware, which as a software developer within a hardware company, it's certainly the model that I follow, right? Hardware is what the computer is and software is what the computer does.

So beyond that, the fact that we now have a specification and standard which is CORBA-compliant, and we have a large number of system vendors -- be it an HP or a Sun or a DEC, as well as independent ORP providers that are delivering that capability-- OpenDoc expands its reach beyond the native desktop environment that it's currently implemented on.

It extends its reach from CORBA to CORBA interoperability, back up to a multitude of server systems that will continue to grow in number as we move out through time. So I see this announcement as providing additional leverage for system houses that are otherwise server-centric in their product deliverables.

UNIDENTIFIED: Yes, Steve's right. I think this actually acts asa springboard for some of the people that have been sitting on the fence and wondering what to do about this space, to use OpenDoc now as a development platform to ensure some independence on the platform they want to develop on. I think you're going to see a lot of software written in this space.

Any other questions? Any questions from anyone on the phone? Or, nothing coming through there? Okay, if there aren't any further questions, let's wrap this up. One thing I do want to do is recognize the fact that there are some ISVs here today that will be showing some demonstrations. Please don't run away, there's booze and food and demonstrations you have to see first.

The ISVs that are here today, some of which will be showing some demonstrations, they will be in the room right through those closed doors which will become open shortly, or you can go right out through the door here where the reception will be. There will be couple of demonstrations right here on the monitors from Apple and IBM that you saw earlier.

I'd like to recognize who those ISVs are in case you didn't know. Bear River is here, Secant, 6 Prime, Part Merchant and Voice Pilot Technology are here today to show you some of the software they've written with OpenDoc.

So I want to thank you all very much. I hope the message has gotten through: distributed objects...distributed objects. C'mon, dream about it. It's where it's going to go. And please join us and talk to us out on the reception area. And thank you all very much for coming. See you next time.

[APPLAUSE! [END OF SESSION! ---------------------------------------------------------------------------- [ Club Home Page | Feedback ] ---------------------------------------------------------------------------- [ Solution Developer Programs | IBM Objects ] ---------------------------------------------------------------------------- [ IBM home page | Order | Search | Contact IBM | Help | (C) | (TM) ] ---------------------------------------------------------------------------- © 1995 by IBM Corporation. All rights reserved. All trademarks are the property of their respective owners.