International Standards Organization Members Approve Sun's PAS Application

Sun Microsystems' Press Teleconference

November 17, 1997 -- What follows is a transcript of a teleconference held after Sun's application to become a publicly available Specification submitter to ISO was approved. Speaking on the call are Dr. Alan Baratz, president of Sun's JavaSoft division and Dr. Jim Mitchell, vice president for technology and architecture of Sun's JavaSoft division.

The official tally:

YES: Australia, Austria, Belgium, Brazil, Canada, Denmark, Finland, France, Germany, Hungary, Ireland, Japan, Korea, The Netherlands, New Zealand, Norway, Romania, Slovenia, Sweden, U.K.
NO: U.S., China
ABSTAIN: Italy, Switzerland.
 

BARATZ: Good morning and thank you to all of you for being with us here today, especially on such short notice. Well, it's been eight long months but it is indeed true that the member countries of ISO's Joint Technical Committee 1 have approved Sun's application to become a publicly available specification submitter. We won our bid with 20 out of 25 countries supporting our application. An overwhelming majority. Of the five that did not, one of them, Russia, was late in submitting their vote and so it was not cast. Switzerland and Italy abstained and there were only two countries who opposed our application, the U.S. and China.

Today is an important milestone for the Java platform. These 20 countries around the world have cast votes in favor of the open, collaborative process that Sun has used to evolve the Java platform. With their votes, technology leaders around the world are saying Yes. Sun is setting a new standard for openness and collaboration in the industry. Yes. Sun's open processes have created a Java platform that is a de facto global standard, and yes, the global community recognizes the value of the Java specifications and is eager to endorse them as standards.

This outcome is not a surprise to us. It is clear evidence that developers, corporations and technologists the world over recognize that the Java platform offers a unique set of benefits that no other technology can provide.

The Java platform has become the software industry's genetic code for two reasons. First, the Java platform delivers value that no other technology can. From Write Once, Run Anywhere to Safe Network Delivery to SmartCard to Supercomputer scalability. And second, the industry plays a vital role in the development of this platform and the process for evolving the Java platform is open and collaborative.

Today's vote also validates the cornerstone of our business model. Keep the standards open and compete on the implementations. We take this vote of confidence extremely seriously. Our commitment to stewards of the Java platform is real. We've made that commitment to 116 Java licensees, to 700,000 developers and to millions of users of Java-based technologies. Sun will continue to propel the Java platform forward in exactly the same way as we have been and that our PAS application defines. We will not change our model. In fact, by virtue of our acceptance as a PAS submitter, we formally committed to the worldwide technology community that we will always define the Java platform in the way that we've done so successfully to date.

We'll work equally hard to use the Java brand to ensure that the Java platform always means Write Once, Run Anywhere to Safe Network Delivery to SmartCard to Supercomputer scalability.

If you talk to anyone in the Java industry you will hear a very consistent story. The entire industry has embraced the Java platform and is committed to what it stands for. The industry trusts Sun to do the right thing and that's exactly what we're doing. In fact, the technology industry has undergone a significant change in the last few years. There has been a real shift in how technology companies treat their customers and developers.

Companies are realizing today that walking people to proprietary technologies doesn't work anymore, especially in the network sector world. The web demands open standards, like the Java standard. The web demands innovation, from not one company but from many. The web demands choice. That's why you see so much focus on open standards. That's why the world voted to approve Sun's application to deliver Java specifications into the standards process. The world has changed. There's no going back. Companies who understand open standards will succeed. Companies who don't will fail.

We are indeed pleased to see this vote today. It is a great endorsement of our vision of open networked computing. It is a great vote of confidence in the entire Java industry. And now I would like to ask Jim Mitchell to talk about the progress which led to today's vote and what we can expect moving forward. Jim.

MITCHELL: Thanks Alan. Just to reiterate. We're very pleased that the national bodies of JTC-1 have given their approval for Sun to proceed towards standardizing the Java specifications and we're grateful to each member country's standards body for their careful consideration of our application over the last eight months.

I'd like to start by clarifying what winning ISO approval as a publicly available specifications submitter really means. The PAS designation is awarded by ISO to consortia and individual companies allowing them to submit information technology specifications that have been widely accepted around the world for consideration as international standards. Consider it a fast track process to have de facto standards submitted for consideration as de jure standards. ISO, the international organization for standards, is the worldwide non-governmental federation of national standards bodies from over 100 countries. By winning PAS designation Sun now has permission to submit to ISO the specifications that underlie our Java technologies for consideration as an international standard.

Why is it important to have an international officially recognized standard? Well, we believe that this is the right thing to do for everyone in the industry. De facto standards are fine and they're what much of the industry relies on today. We believe, however, that the industry is best served if a technology of this importance is validated by an international standards organization of the stature of ISO and IEC. Many governments, universities and other institutions require use of standard technologies in their procurement policies and requests. ISO approval will help to broaden the market for Java platform providers and software developers.

Having said that, it is also clear to all of us here that the Java platform is already a de facto standard which much of the industry relies on. The unique value of the Java technology and the open processes by which the Java platform has evolved can be created with this widespread adoption.

Sun's application to become a PAS submitter broke new ground and as a result generated a fair amount of controversy. In submitting the PAS application we were the first company to apply to use this new way of making standards relevant in the radically fast paced technology environment in which we all work. As the first corporate PAS applicant Sun wanted to set the bar for other companies who will follow and we think we've done that. Our commitment to open specifications and open processes for developing them is an excellent model for JTC-1 when considering other applicants who may follow.

Indeed, the mechanisms by which the Java platform developed are a new model for new times. We started with a technology that delivers value which no other technology can. We developed this technology based on principles of speed, openness and industry-wide participation. Although Sun invented the Java technology, there is major intellectual property in the Java platform, not only from Sun, but also from IBM, Netscape, Borland, Intel, Microsoft, Novell, Oracle, and many, many others. The strong result in favor of Sun's application shows that the worldwide standards bodies understand and approve of Sun's open development process for defining the Java platform and that our stewardship has been key to its success. We can move ahead with our process with confidence knowing that there is a well defined route, the PAS process, for creating an international standard based on the Java platform specifications. This vote is a vote of confidence from an esteemed international community of technologists and scholars who are eager to see the Java platform become an officially endorsed standard, and of course, we are very pleased with the outcome. We're all ready now to take your questions.

OPERATOR: Maria Georgianis from Dow Jones.

GEORGIANIS: Good afternoon. I just wanted to clarify for myself that Sun has permission to submit to the ISO the specification for the Java platform that would put it on track to be accepted as an official standard. So I want to know what the next steps are. How soon will you submit these specifications? When do you anticipate that it would be approved as an official standard and how many member countries did you need to vote on that, to approve it?

MITCHELL: We were told by ISO that what we needed for the approval was a simple majority. Of course we got much more than that. We got a 90% majority of yes's over no's. But their rule for this vote was a simple majority. In terms of when we will submit something, we really wanted to wait and see whether or not we would win our bid to become a PAS submitter, so it's a little premature for us to say precisely when. There is real work on our part to take the Java platform specifications and transpose them--that's the official standards term--to turn them into the format and everything that ISO wants. So we'll begin to look at that but we do not have a timetable for it right now.

OPERATOR: Dave Akin, Hamilton Spectator.

AKIN: Thank you. I'm calling from Canada and of course Canada was one of the ones who voted yes. One of their concerns was whether or not Sun understood the concept of being a PAS. Specifically they took that to mean is, the Java standard and the Java trademark which you own will be two separate things, and right now I assume at the early part they will be identical. If somebody has the Java trademark, he will have the Java standard. But conceivably I could submit something to the ISO to improve the Java standard and all it would take would be an ISO vote, and it could be against Sun's wishes, that something is added to the standard. So there is a possibility that at some point down the road the trademark, which you own, could be different from a standard. I'm wondering if you've thought about that, if that concerns you or what your reaction to that might be.

MITCHELL: First of all, I've got to give you a little bit of process about what happens when we actually do a PAS submission now. When we do that we actually have to enter into a negotiation with JTC-1 about how that submission is going to be handled, that includes things like maintenance, how it changes and evolves and so on. Of course, we haven't done that yet so I can't tell you precisely what that is going to be. We did get a lot of feedback about that part of the process from the national bodies. We've taken that into account in our responses to them, and I think that's part of the reason why there was this very dramatic outcome at the end. It is not guaranteed because the end result of that negotiation with JTC-1 exactly how changes to the standard will be made after that. Our primary concern, our overriding primary concern is that we have to preserve "Write Once, Run Anywhere". We will not do something that's going to destroy that, whether for another company or for a standards body. It's very important for the world that "Write Once, Run Anywhere" be preserved. I think ISO understands that, so we'll be able to work together with them and figure out a way so that we don't have that kind of bifurcation going on because that would be a bad thing to do and lower the value of both the Java platform and the Java brand. I'm sure ISO doesn't want to do that.

AKIN: Could you comment on the significance of the fact that the U.S. of all countries -- home to not only your company but many other global leaders in the IT business -- did not endorse Sun's application and become a PAS submitter?

MITCHELL: There are a couple of responses to that. We certainly would have preferred a "yes" vote from the U.S. TAG. We worked hard to get it. We actually got a 60% majority, we just didn't get 67%. They weren't actually against. They didn't vote strongly enough to overturn their previous "no" vote. I would have liked those extra couple of votes to make it that way. But in JTC-1 each national body has one vote and I think that's entirely appropriate. It isn't just manufacturers who should be represented on standards bodies but consumers as well. The whole world is a consumer of Java technologies. We're actually happy with the overwhelming support. We had fairly strong support in the U.S. vote in spite of the fact we held the meeting right in Microsoft headquarters. It was a simple majority, not a two thirds majority. But even in the U.S. we have a majority of the people on the standards organization who thought that approving us was the right thing to do.

OPERATOR: Michael Moeller, PC Week.

MOELLER: A couple of quick questions. You've won the right to become a Publicly Available Specifications submitter, but what does that really mean? Does that mean that all standards or specs or pieces that make up the Java platform have to go through Sun before they can be submitted to ISO? Likewise, what are logically the first pieces that you're planning to put forth into the ISO process? Also, what happens if things change and it gets batted about much like C++ and either innovations get stalled because of the standards process or things get changed that benefit one platform over another?

BARATZ: Mike I just want to answer one piece of one of the questions you asked and then I'm going to turn it over to Jim to answer the remaining question. One thing to keep in mind is the process that Sun defined for evolving the Java platform. The open industry participative process that we have defined and that we have used for the last year and a half is not changing. So we're not talking about the evolution of the Java platform now being done by ISO, ending up in committee that will slow down the definition and the evolution or cause compromise. The Java platform will continue to evolve in exactly the same way that it has to this point in time. There is a process that we have defined which involves key experts from different countries to write an initial spec, review by licensees and then open review on the web where at each stage we very seriously consider the comment and modify the spec as appropriate based on those comments. Submission then to ISO is a next step beyond the process that we have been running and will continue to run for the evolution of the Java platform. With that, I'll turn it over to Jim.

MITCHELL: That is exactly right. One of the reasons, Mike, for the whole setting up of the PAS process was so that de facto standards could get taken and become de jure standards without going through the working group process that usually changes and massages things. That, of course, was one of the attractions to us when it was suggested to us by JTC-1 members. We didn't want to turn the Java platform into a standard then on the way have it invalidate all the investments made by the Java industry out there. That would be I think close to criminal. It shouldn't suffer the fate of C++ at all because it's not going to be, if you will, developed in committee. Once you're an approved submitter you take ISO a set of clean specifications and they vote, and there is ballot resolution. They certainly can comment on it but it doesn't go into a working group, into committee where it gets massaged over a long period of time. It is a voting and ballot resolution process and not a working group process.

MOELLER: So there is really the working group and the Sun partner companies that are working to define a specification. If a spec is submitted to ISO and it is voted down, you then would have the opportunity to change it based on any feedback made at that voting process and then resubmit it for another vote?

MITCHELL: You're close to right. Let me modify two things you said. One is we work on the specifications not with just our licensees but in fact with any developer on the web and they have made fundamental architectural changes when they've made their comments. It's actually much broader than that. It's broader than any standards body. The second part of that is that the voting process is always a two step process. You put it out in ISO and you typically get back comments and that is part of what's called ballot resolution. You pay attention to those comments, make changes or whatever and then go to a final vote. Just like we did as applicants to be approved submitters.

MOELLER: What the first pieces of technology that you're considering putting into the ISO process now?

MITCHELL: We've been very clear about that. The specifications for the core Java platform because that's what preserves "Write Once, Run Anywhere". It's the language, the virtual machine and the class library specifications. And it's specifications, not code. It has nothing to do with code, only English words.

OPERATOR: David Carr, Web Week.

CARR: My understanding is that the votes you got that were "yes" votes still had comments attached to them. A number of the delegations still had concerns that they expressed along the way. Is that what's giving rise to the negotiation that Jim Mitchell referred to?

MITCHELL: You're right that a number of the "yes" votes came with comment. As part of actually doing a PAS submission--this is in the JTC-1 description of the process -- one is supposed to enter into a negotiation with JTC-1 about a variety of issues. Maintenance is one of them. Most of the comments as I understand it--I actually haven't seen them yet--are still about the maintenance issue. That was always supposed to be worked when we actually did a submission. So what will happen is when we do a submission, those comments will be essentially the first issues on the table. But that is a standard part of the PAS process. You come up with what's called an explanatory report and it's really what's voted on is the explanatory report plus a submission.

BARATZ: I think it's also fair to say Jim, that comments associated with the "yes" vote are not relevant with respect to accepting or rejecting Sun as a PAS submitter. The JTC-1 process doesn't recognize comments in this second round of voting. It's simply a "yes" with a 20 vote out of 25 majority and at this point we are now a recognized PAS submitter. No constraints, no more negotiation. We are now entering into the normal process that any PAS submitter will enter into with respect to making an actual submission.

CARR: You don't see any danger that somehow you would be unable to come to an agreement on management and would either wind up with Sun walking away and deciding never to pursue this or JTC-1 saying that what you're proposing is not acceptable?

MITCHELL: You don't have a negotiation if you know the answer ahead of time. So it's a negotiation. I'm pretty confident about the negotiation but you can't tell what the result is until you have it.

OPERATOR: Torsten Busse, IDG News Service.

BUUSE: Since China and the US voted against the PAS status, I wonder what type of policy you'll run into when the international members will once again vote on the Java standard itself?

BARATZ: Well, two voted "no" and 20 voted "yes" so I think that's the right way to look at it. The process will run in a very similar fashion when it comes time to vote on the specs themselves and rather than dwelling on the 10% that voted "no", I would dwell on the 90% that voted "yes."

OPERATOR: Tim Clark from C/Net.

CLARK: I'm still a bit confused about the process from here forward. It sounds as though you've been accepted as a PAS submitter. There are 13 people who have voted "yes" but have some objections. I think the answer was you deal with those objections at the time that you submit the entire specification. Is that right? What happens after the specification is submitted? Is there a second round of voting or what?

BARATZ: The way you've characterized it is not correct. Sun has now been accepted as a submitter of Publicly Available Specifications for the Java platform. Period. The comments that got associated with the "yes" votes are not relevant for the vote that was just taken and so they're basically removed from the vote. The count is tallied 20 "yes", two "no". Sun is accepted as a submitter of Publicly Available Specifications for the Java platform. The next step is for Sun as a recognized PAS submitter to make a submission. The defined ISO process for a PAS submitter to make a submission includes a identification of the maintenance process. At that point there will likely be a discussion with ISO around that maintenance process. We are guessing that there will be a discussion and negotiation by virtue of the some of the comments associated with the "yes" votes at this point, but those comments have nothing to do with recognizing Sun as a PAS submitter for the Java platform specifications. They are just something that ISO is aware of and to the extent that they bear upon the normal process for now making a submission, they will come back and be a part of the discussion at that point in time.

CLARK: Can you tell me what the objections were? The press release from the JTC-1 says you have received the comments.

MITCHELL: We haven't yet received the comments. They said they sent them. They didn't say we received them. We should get them today. I assume they're going to be sent to us by e-mail but I certainly have not seen them. If I did I would be happy to talk about them, but I just haven't.

OPERATOR: Anne Thomas from Patricia Seybold Group.

THOMAS: Congratulations, I'm very pleased. My question is which version of the Java platform specifications you plan to submit?

MITCHELL: The one we will submit will depend upon when we can get it done and what is the de facto standard at the time. We don't want to submit something that is old and behind its time and we don't want to submit something that hasn't arrived yet. It's somewhat tied up with timing of actually doing the submission. I personally want it to be something that is mature in the sense of having been used a bit. I just think that's responsible.

THOMAS: Do you figure this is going to be three months out, six months out, twelve months out?

MITCHELL: It certainly is not going to be three months out. It's got to be longer than that. There's real work to be done. If you add up the pages and the Java specifications that I've mentioned so far, the books themselves, three of them together, are about 5 inches thick. That's a lot of work. This isn't like easy prose right. This is hardcore technical stuff that has to be absolutely correct. There is some real work to be done there. It's certainly not going to be in the three month timeframe. It's going to be beyond that.

OPERATOR: Norri Kageki from Nikkei Business.

KAGEKI: I have a question for Jim Mitchell. When you answered the first question you said you're in the "wait and see" mode. What do you need to wait for?

MITCHELL: What I said was we had not started the work to transpose the standards because we didn't yet know what the outcome of the vote would be and it didn't seem to be a good idea to be starting the work if we were going to get a "no" vote. That's all I meant. We are not still in a "wait and see" mode but we were before the final results were in.

KAGEKI: Are you going to hire new professional people to deal with this standardization matter or are you going through reorganization?

MITCHELL: Nothing like that. Of course there are versions of these specifications that already exist and they're actually very high quality. Their out as Addison-Wesley books from the source. The Java language specification has been remarked on in the computer science community as one of the very best language specifications anyone has seen. That's got James Gosling, William Joy and Guy Steele as authors, three very strong computer scientists and one the inventor of the language, so we don't need to hire some new people to get those specifications updated and transposed.

KAGEKI: But I think you need a completely new group of people to deal with ISO and with all these hundreds of papers.

MITCHELL: Sun has been involved in standards processes for quite a long time and we have a set of people inside of Sun already who are representatives on the various standards bodies around the world. This is not our first attempt to be involved with standards at all. We've actually been doing it for a very long time so we already have those people.

OPERATOR: Jeff Kinz from International Data Corporation.

KINZ: I've been looking over some of the market data and predictions. The U.S. holds currently, as of 1996, about 35% of the entire worldwide market for software. China, of course, is a market that is going to expand rapidly in the next decade. We see a situation here where a significant portion of the consumer market appears to be positioned against accepting the Java specifications coming from Sun. Can you comment on that? Is that a true perception?

BARATZ: I think what we see here is a situation where a small number of companies in the U.S., one of them software and three of them hardware, simply lobbied very hard -- including putting pressure on some of their partners -- to just barely escape a 2/3rds majority "yes" vote in the U.S. It was a two company difference in the U.S. between the 2/3rds majority "yes" versus the simple majority "yes" that the U.S. did provide. So, no, I do not think that the U.S. vote in any way, shape or form reflects the fact that the majority of software companies in the world have a problem with this submission. Quite the contrary. I think that if you go and look at the companies that have voted to support this, there is overwhelming support among the software company community that this is the right answer for the industry.

KINZ: Following up on that, how many companies voted "yes" and how many would have been needed to make a 2/3rds majority?

BARATZ: In the U.S. the numbers were: 15 companies voted "yes," 10 companies voted "no" and 6 companies abstained. We had a simple majority and if two of the "no" votes had been "yes" it would have been the 2/3rds majority.

KINZ: So despite the results of the votes most of the member companies in the U.S. were in favor of the Java standard coming from Sun?

BARATZ: Yes, this is correct.

OPERATOR: Larry Perlstein, Dataquest.

PERLSTEIN: We touched on this earlier, but I was wondering, once the specifications are approved, how do you foresee the timing of the inclusion of new functionality into your product versus its adoption into the spec itself?

MITCHELL: Let me take a crack at that, first. Of course international standards typically always lag implementation because they require more agreement and so on. That's been true for many years. There have been implementations of C++ out before there ever was a standard. Indeed, there is still isn't an approved standard for C++ but there are products out there. It always leaves the problem that if the standard ends up being a little different than the products, the products that want to be able to claim that they adhere to the standard, that they conform to it, usually have to go through a little extra release cycle than to conform to the standard. That's always a problem. It's no different in this case. And indeed it's helped by the fact that the PAS process allows one to get into the standard what is actually being shipped by the platform providers. That's a big help. It means that the changes for an initial submission should be quite a bit less and it should be easy to get everybody lined up so that folks can be shipping things that are ISO conformant.

PERLSTEIN: Is it fair to say that you accept that your implementation as well as your licensee's implementations would all be somewhat advanced from the specification?

MITCHELL: Well at the moment our licensees pretty much are the computer industry. There is over 120 licensees and it's all the major high volume and medium volume platform providers. The good news here is that it would be more widely adopted and more implementations done than almost any other standard I can think of in ISO.

PERLSTEIN: How does the specification really benefit?

MITCHELL: The spec benefits the Java industry where there are countries, and there are, and companies who for one reason or another prefer or need to require that technologies are based on international standards. It helps in those marketplaces. It also helps that an international standard is inherently more stable, that advance more slowly, than the industry typically does, so people can count on it for a longer period of time.

OPERATOR: Debbie Gage, Smart Reseller.

GAGE: If there is a lag time between actually submitting the specification and working on it, does that give Sun extra power in how the spec should be determined? Will your licensees listen more carefully or feel obligated to design Java products the way Sun would like thinking they might be in danger of deviating from the standard, or is that not an issue?

BARATZ: The thing to keep in mind is that as a submitter of Publicly Available Specifications for the Java platform, Sun is responsible for defining how the spec gets created and what the spec ends up being. That's what it means to be the submitter of Publicly Available Specifications. It is not going into a separate working group for further submission to ISO but rather it is through the process that Sun has defined that the spec will be created and will then be submitted to ISO.

Now, we have gone on record as stating that we are committed and will continue to be committed to evolving that spec through a very open industry participative process. What this means is that for any element of the Java technology, the process that we run will have experts from several different companies involved in defining the specification. It will then go out to review, including worldwide review through our website to anybody who cares to comment, and we will seriously consider those comments and make changes as appropriate based on those comments. As a result, there isn't going to be any one company that will have an advantage with respect to the specification because there isn't any one company that is going to be defining the specifications for any one part of the Java platform. However, there is one company that is responsible for the process for defining the specification. That is Sun and Sun has committed that the process will be open and industry participative, involving many companies from around the world and an open review process.

OPERATOR: Maria Georgianis.

GEORGIANIS: Some of these are very basic. Are you the only party that can submit the Java platform specifications? Secondly, after you submit it as a specification, I don't understand the review process, if you need the simple 2/3rds majority to accept that. Third, why is your specification in the past described as not open?

MITCHELL: I'll try all of these, in reverse order. I think the reason that our process was described as not open was that there are companies that didn't want us to be approved as submitters. That's rhetoric as opposed to truth.

We have vast amounts of evidence about the input to this process and fundamental architectural changes being made due to review by developers on the web and so on. It's just not true that it's not open. In fact, as a result of our doing a PAS submission, a number of companies and some standards bodies have requested copies of our white paper describing our process because they want to adopt certain of its characteristics. There's truth -- votes by emulation that it's open.

The review process, when we submit something, is actually a process called ballot resolution. There are two things that happen. We get into a negotiation with JTC-1 around a number of issues. It's a limited number of issues: maintenance, intellectual property rights are essentially what's involved. Intellectual property rights is just so they have copyright. We have to negotiate that with them and produce an explanatory report and that explanatory report says how maintenance will be handled, what about intellectual property rights, and so on. That, along with the specification, gets voted on. In this two step ballot resolution process that allows comments after the first vote and then you deal with those comments and then go to a final vote to get a simple thumbs up or thumbs down. Very much like what happened as we applied to be a submitter.

We are so far the only company that's been approved as a PAS submitter. The process was always intended to include companies but part of the pioneering here, both on our behalf and on JTC-1's part, is that we're the first company to actually apply and to be approved. We're approved as a submitter. No one else has been and so in terms of the Java specifications, which at the moment have our copyright on it, we're the ones who are going to do that submission.

BARATZ: I'd like to add one thing. I think that if you were to compare our process one to one against say the process that Microsoft uses to evolve its technology and ask which of those processes is open or at least more open, I think you'd find a world of difference between the open industry participative process that Sun has been using for the evolution of the Java platform specifications and the closed proprietary process that Microsoft has been using and other companies have been using to evolve their technology. In fact, I would call on Microsoft to rebuild their process identical to the one we use and apply to ISO to be a submitter of Publicly Available Specifications for Windows. I would love for Windows to be an ISO standard. I would love for Windows to evolve in the way that the Java platform evolved. I would love for Microsoft to be half as open in their process as we are in the evolution of the Java technology.

OPERATOR: Dave Akin.

AKIN: I just want to get back to the issue of shareholder value for Sun shareholders. Set me on the right track here. Shareholders will find value in this process essentially because as you mentioned, if the Java platform becomes a standard then that opens up a bunch of new markets for developers writing software using the Java platform specifications to sell the public sector governments, such signatory treaties like the WTO and NAFTA. That's where Sun is going to derive its shareholder value out of this arduous and complex process?

BARATZ: I think there are a couple of ways to think about the benefit or the value from Sun having been accepted as a submitter of publicly available specifications. First of all, the success of the Java platform itself is obviously very important to Sun Microsystems as well as to many other companies around the world. The whole notion of "Write Once, Run Anywhere" and network safety enables a whole new generation of network-based computing which yields new products and services for many companies around the world. Moreover, the notion of the Java platform's "Write Once, Run Anywhere" network safety characteristic is releveling the playing field in the software industry. Now you can go buy microprocessors or operating systems based on the characteristics of those products that are important. Things like price performance, reliability, configurability, scalability, rather than by the application lock-in that exists today in the industry. This has enormous value across the industry for all but the small number of monopolies that exist today. The fact is that the Java platform provides a common programming environment across all operating systems and microprocessors, so that application developers can build a base of applications that will run on all operating systems and microprocessors, so that when you go buy an operating system or microprocessor, you don't buy it based on the applications that run on it. You buy it based on what's good about that operating system or microprocessor -- that is good for you as an end user, that's good for the industry building applications, that's good for the industry building operating systems and underlying system software. It's a releveling of the playing field that the Java platform enables.

That having been said, what has happened today is that ISO has essentially said the process that Sun has defined for evolving the Java platform,for ensuring that "Write Once, Run Anywhere" is always preserved; for ensuring that network safety is always preserved; to ensure that scalability from super computers to smart card is always preserved. We think that that's the right process evolving the technology and so we are not going to insist that there be another level of working group that gets involved in trying to massage or modify this technology and potentially erode the value proposition of the Java platform. The first element of value that comes from ISO accepting our application to be a submitter of Publicly Available Specifications is commitment to the process we've defined that is designed to be open and industry participative, and to preserve the real value proposition of the Java platform: "Write Once, Run Anywhere", network safety, smart card to super computer scalability. This has enormous value to Sun and enormous value to almost every company within the industry and end users as well.

The second element of value that comes out of this is simply the notion that there are a number of government institutions or universities or other entities that are uncomfortable buying a product unless it's based on ISO-standard technology. As a result, this enables Java technology to be a part of the products sold to those government institutions or universities that are committed to buying product that includes only ISO-standard technology. So, the value here today is in commitment to the process that yields ongoing preservation of the Java platform's value proposition and then the path that allows us to turn the technology that comes out of that process into technology that is ISO standard so that entities that buy only ISO standards can benefit from this.