The fifth chapter reversal miracle - the history of the battle of Orland JBuilder "No jbuilder, Borland is impossible to have today's Rongjing!" Java's rapid rise and success no one is expected, even for Sun himself seems to be a Extremely unexpected, but the successful person must be decisive and act quickly. When the Sun detects the bright future of Java, we will start vigorously sell Java. Mr. McNealy, Mr. McNealy has suffered from the opportunity to fight directly and Microsoft. This seems to find an opportunity in Java. Of course, it is more important that Sun's next series of action has proven to be correct and successful. These actions include cooperatives with various manufacturers; Publish a series of best-selling and successful Java books with Addison-Wesley; publish all the media related articles related to Java, column, etc., quickly cultivate the foundation of Java users, Attractors' interest in Java. This is completely Microsoft has always been uneven, and there is no way to attack, SUN also has to play the fullestity, and "the way is still in the way." More importantly, McNealy has put a large number of R & D resources immediately, constantly improving Java, and finally makes Java began to show the front of the industry since 1995, and quickly became the industry's focus, and the biggest scale in the history of PC fight against Microsoft battle Also change the habit and direction of many software development. Of course, for Borland, Java's development history is also a battle against the waves, and Borland has never experienced a large-scale group military mix. For Borland, things are not so smooth. In 1995, when Java started to take off, Borland did not expect Java's growth rate so fast. Borland started Java as a C / C extension, so P1UG-IN to support Java is only added in Borland C / C 5.0. However, Borland quickly found that things were not so simple, because in addition to Java's PLUG-IN reactions, Symantec quickly found a new stage in Java development tools, and developed fairly quickly. Before Microsoft's attitude towards Java is not clear, it is undoubtedly Symantec to occupy a leader, Borland is alert to his own disregard, and the Java General Bay starts Borland behind it behind. How can Borland be counterattable in the next most important development tool war? After the early stage of Java development tools, when Symantec retreated from the C / C development tool market, Eugene Wang is a fairly high-income development tool, and it is obvious that although the C / C market has failed, it can be used. The Java market that is about to thrive will take back to a city, so it immediately leads the original Symantec C / C development team to quickly enter the field of Java development tools. Eugene quickly developed Java development tools as the foundation of Symantec C / C , started to develop Java development tools, which is later famous Visual Cafe. Symantec is almost the first software company involved in Java development tools, using Symantec C / C foundation, so in 1995, when Java got the attention of more people, Symantec was also ready for her first. Java development tools - Visual Cafe. In October 1996, Symantec officially launched Visual Café before JDK 1.1.
Although many people criticized Symantec and then launched after JDK 1.1 to support the latest JDK standards (because there were many issues of JDK 1.0x version at the time), but these criticism did not hinder the success of Visual Café. Since many software workers were eager to invest Java's learning ranks, they immediately got a great success at the market immediately after Symantec launched Visual Cafe. Especially in the Java learning market and education market, Visual Café is almost the status of the first place in the Java development tool, becoming a hot product, and Symantec is also frustrated in C / C development tools. Resentment, once again become a leading manufacturer of the development tool market. Since Microsoft adopts an enemy attitude toward Java, it is almost impossible to launch Java development tools, while Borland is also in the hard work of C / C , has not been confirmed by Java potential. As for another dead endcom, it has been acquired by Sybase, and it is impossible to build a climate in the development tool market. This is a good opportunity to be a good opportunity to solve the Java development tool market for Symantec. The only threat left is the Java development tool to be launched by Sun. But Symantec has grabbed the market, and has become a leader. As long as it is well grasping, it will be able to fight for the Wanted forces and SUN. In this stage of the Java development tool, Symantec seems to be a absolute advantage, but it is a pity that Symantec has also received several mistakes and gradually lost the advantage. First of all, when Visual Café is launched, Eugene Wang left Symantec to do business. This has a considerable impact on Visual Café, because Symantec is able to fight against Microsoft, Borland and Watcom in the C / C development tool market; Eugene Wang has created the first one in Eugene Wang Java development tools Visual Cafe. Symantec should rely on the Eugene Wang's skill in the early Visual Café, but Symantec actually lets Eugene Wang leave, and immediately develop a generals of the development tool. The second error is Symantec to introduce Visual Cafe as soon as possible to seize the market opportunity, so the integrated development environment is written in C / C language. This has caused several disadvantages, one of which is due to the use of C / C to write Visual Form Designer, so the performance of the programmer is designed and the effect of the real Java program is executed. Some differences; second, in order to maintain Java's control component program code and keep the synchronous state of visual formwork in C / C , in the original program code generated by visual form design, some Visual Cafe Control Tag. These control volume labels are not Java program code, just for visualization form designers. If the programmer accidentally modifies or deletes these control volumes, it will cause the Visual Café's visual formator to fail. This is very serious shortcoming, Symantec should improve these problems immediately after Visual Café 1.0, but Symantec seems to be impossible to improve effectively. Of course, the root of the problem is that the integrated development environment of the Visual Café is written in the C / C language. To improve this problem, Symantec must rewrite the integrated development environment using the Java language, which is the strategy that Borland later taken later.
However, it is also very adventive action using Java's integrated development environment. Borland has lameled heavy cost. Symantec did not do this, probably because Java at that time was not mature to the point you can do. But Sun is obviously not so much, it has a hundred times of confidence in Java, and soon the Sun's Java development tool is launched, Java Workshop. When Java successfully captures the developers' heart, the requirements for Java development tools are increasing. Although Symantec has launched Visual Café, many people still hope that Java's authentic manufacturers Sun can launch Java development tools, let all developers who want to learn, developers using Java can use the most standard Java development tools. Of course, many people also hope Sun can use Java to write Java development tools to prove that Java's ability to resist, let people questioning Java capabilities and Microsoft shut up. Of course, Sun is confident that Sun's development tool plan is confident after Java success to meet the needs of the majority of developers. McNealy has attracted the hope of Microsoft sites for many years. It seems that bright future. Shortly afterwards, SUN finally launched the Java Development Tool Sun Workshop. Symantec is very nervous before Java Workshop is about to be launched, as this is related to whether Symantec can stand firm with the Java development tool market. I remember that in Java Workshop, all media, magazines reported significantly Java Workshop, Sun also vigorously promoted and promoted Java Workshop. From the moment of momentum, Java Workshop has the taste of "Qianqiu Wanshi, a Rivers and Lake". Many friends I know are eager to get a set of Java Workshop, I am eager to find a set of Java Workshop, whether it is bought, borrowed, download, etc. However, the ugly daughter-in-law will always see the in-laws. After many people used Java Workshop, they found that they did not only perform slowly like a turtle, and there were a lot of problems, and the general PC development tools were more than 100,000. in. It seems that Java Workshop is only suitable for use on an expensive Sun workstation computer, and at that time, the PC used in most people is running at all, unless they have the resistance of ordinary people. Java Workshop thundered is small, and it is popular in the shortcomings. Many people who have confident in Java have started to question Java to develop complex applications after using Java Workshop. Is it only suitable for writing Applet? Is it only suitable for use on Sun's workstation and computers? Although Sun is still working hard to launch the Java Workshop 2.x, I hope to wash the java workshop 1.x, but still unable to save the confidence of Java developers. As for those who are interested in Java, they use Symantec's Visual Cafe to further expand the market share, but also make Symantec aesthetic pill. Of course, many Borland supporters have begun to intensify and require Borland to launch the best Java development tools. After SUN is defeated by the Java Development Tool market, the PC development tool market is not the same as the Solaris development tool market. SUN on Solaris is a unique, but in the PC market, it is a hundred families, and the competitors are more powerful than one.
Sun does not understand the characteristics of the PC development tool market, thinking that it is a big mistake to pass the Java authentic sign, and it is a big mistake, and at the time by Microsoft, I don't know how to develop software, this is because Sun often laughs Microsoft does not understand development operation. The system, it seems that Sun is not 50 steps to laugh. SUN gave a gray face after the Java Development Tool Market, had to develop Java language and JDK library, and started to develop Java component technology after the Java language was more mature, so I opened a later and Borland cooperation to develop Java. The BEAN's functional specifications, and then borland's specifications, and finally a strong interest in Borland's JBuilder, and even want to merge Borland. Of course, this is because Borland's highly technically high technologies, so that Sun is caused by the cause of affirmation to the concession. Borland's Java hard struggle "Don't be so smooth", Borland's R & D director said, and full of anxiety. When Borland alerts to Java's potential, Visual Café has been successfully listed, and Sun is also ready to launch Java development tools. At that time, Borland was gradually lost from the C / C market, and the financial deficit was also started. The entire company is in the case of a mess, it seems that there is no additional resource to invest in Java research and development. In the case of starting behind, in the case of less food, Borland seems to be willing to lose entered the Java market. Fortunately, Delphi is red, let Borland earn a vote, but also stabilizes the military. Delphi's resource injected for Borland also allows Borland to activate the Java R & D team. Although Borland has been behind, Borland knows that it is never lost this market, because Java's market does not have Microsoft's oligomatics, Borland has a better success in the Java market than Borland C / C , Delphi. In addition, Borland needs to explore other income sources outside of Delphi's product line, otherwise it will only rely on Delphi products, and the company still can't grow more, and compete with other software companies. In 1994, between 1995, Borland officially established the Java Research Group to start developing Java technology, preparing to develop Java development tools. The internal R & D name of this Java development tool is Latté. At the beginning, the R & D resources of the Latté group were not enough, because the Borland at the time was in the wind and rain, and enough resources were unable to inject enough resources to the Latté group. Therefore, the initial progress of Latté began development is not smooth, and the progress is slow. The Latté team has been sufficient for the Latté team after Borland relying on the Delphi Bath, and the progress of R & D began to accelerate. However, compared with the competitors, Borland is indeed quite backward in Java, almost running in the final entrant. Fortunately, the battle of Java development tools seems to be a long marathon, except for the beginning of the performance, more important than anyone can support it. In fact, how to defeat the defeat in the Java competition, one by one, and then become the process of the Java Development Tool King, and the competition strategy used by the JBuilder team is more worthy of our taste and learning. According to my personal vision, I have experienced several different stages in the process of Borland Development Java Development Tools, and every stage has very intense competition, with successful and losers. Only some of the losers still insist on competition, and some have disappeared with the wind.
JBuilder can ultimately become a king, except because it is more than healing, Borland does not quit the Java market, but Borland has a critical decision in developing JBuilder 3, and each version has a clear goal after JBuilder 3. I finally slowly became the first leader in the market after JBuilder 4. Of course, this long-year-old battle process is very hard, but this history is the history of the entire Java development tool. Section 1 - Java Jit Compiler's War Borland may not start developing Java technology manufacturers, but it is indisputable to other competitors. Borland has not had the case of Latté, and the first step in the Java competition is to start from the Borland tradition, which is starting from the Java Jit compiler. However, because Borland has not had good mastery for Java technology, it is started with PASCAL's ancestors Dr. Niklaus Worth, by Dr. Niklaus Worth, and his students develop Java Jit compiler for Borland, and Borland itself The Latté team developed Latté other features in parallel. Since Java has gradually popped up in the campus, it has attracted many academic research interests, Dr. Niklaus Worth, and his students have been put into Java-related research. Therefore, when Borland is looking for the door, naturally, it will be taken. Borland has shortened development time, while Dr. Niklaus Worth Research Group is happy to sponsorship costs. The first work of the Dr. Niklaus Worth Research Group is the Java Jit compiler launched around 1997. This JIT compiler developed by the Dr. Niklaus Worth Research Group allows the compiled Java Bytecode to perform speed than the sun of Sun's Java. The compiler and the Symantec's JIT compiler quickly several times. Borland announced that this JIT compiler immediately shocked the Java boundary, because the slow Java execution speed is that all people who use Java hopes to improve significantly. The Java Jit compiler launched by Borland seems to have seen the future hope for all Java developers. Although it is strictly that even if the Java program is compiled using Borland's latest JIT compiler, the execution speed is still very "turtle", but for the use of Java to learn or write, do some small applets still It is very easy to use. Therefore, when Borland launched this JIT compiler, it immediately responded to the popularity of Borland in the Java world. All Java developers have also begun to focus on Borland as a serious competitor. Otherwise, in addition to the momentum of Borland, Borland has almost all. Borland's Virgin Java Jit compiler is red, immediately attracted the attention of the browser's domineering Netscape. Since Netscape strongly supports Java to compete with Microsoft, it is necessary to have quality Java Jit compilers in Netscape to smooth and quickly perform Java Applets, increase Netscape's competitiveness and attraction, highlighting with Microsoft. The difference between IE. Soon after Netscape found Borland, I hope to come with Borland's Java Jit compiler in Netscape. For Borland, this is another chance of chance. Because this not only proves that Borland's efforts in Java technology, more importantly, Netscape is a software company that is not always in the world, millions of users around the world.
This means that once Netscape built Borland's Java Jit compiler, Borland will immediately have millions of Latté potential users in the world. For Borland, it is better to have a good condition. Therefore, Borland immediately agreed to Netscape proposal to let Netscape match the Borland's Java Jit compiler. But this move also immediately pulled a whole body, which in turn caused a big blend of Java Jit compiler. After Netscape and Borland reached an agreement and started shipping, it caused Symantec's worries and dissatisfaction. Because Symantec is the boss of the Java development tool, but Borland's Java development tool has not been launched, but Netscape runs to use Borland's Java Jit compiler, which is not letting the world know the strength of Borland and let Symantec face Is it? For colorful faces and avoid the market that lost Java development tools, Symantec decided to start counterattack. Symantec immediately focuses on the development of the Java Jit compiler, develops a fast Symantec JIT compiler than the Borland Java Jit compiler, and ready to develop a Java compiler that compiles Java Bytecode to native Windows program code. Somantec also announced a new Java Jit compiler after the Borland Java Jit compiler. Symantec's Java Jit compiler is more efficient than the Borland Java Jit compiler, compiled Java Bytecode execution efficiency is 2 ~ 3 times faster than Borland. After the Symantec Java Jit compiler announced, it was reversed. Orland's face was light. I just got a good level of cooperation with Netscape. I didn't think that the efficiency of the throne was still kicked by Symantec. How can I treat Netscape? Therefore, Borland immediately improved the research and development of the JIT compiler, and strive to exceed Symantec again. Sure enough, Borland's efforts did not have a white feet, soon, Borland's JIT compiler broke the efficiency record set by Symantec JIT compiler. Since this Borland and Symantec, the "Racing" competition of the Java Jit compiler is launched, and the other party is constantly trying to defeat each other. Due to the JIT competition of Borland and Symantec, of course, more important reason is that Java execution speed is too slow at the time, causing IBM, Microsoft, and Sun in Java compiler. Symantec is worthy of the king of Java development tools, after making a few JIT compilers several times in Borland, they have gradually accounted for the wind. The Java Jit compiler developed by the Dr. Niklaus Worth Research Group is also no longer the opponent of Symantec. To this Borland decided to recover the technology of the Java compiler and started its own development. BORLAND finds that there is no big use with Symantec in Java Jit compiler. It is a matter of urgent to launch its Java development tool. Therefore, Borland starts with Symantec in the Java Jit compiler competition to promote Latté. Of course, Borland exits the first phase of the JIT compiler that is shortly after Netscape no longer uses Borland's Java Jit compiler, change to the Java Jit compiler using Symantec. To this, Symantec finally won the victory of the first phase of the JIT compiler, saving the face of the Java development tool first manufacturer. But is SYMANTEC really win? That can't be asserted because the JIT compiler war has just begun.
After Symantec's Java Jit compiler, Symantec was moved to Sun, and I hope Sun can also use Symantec's Java Jit compiler to push Symantec to Java core technology leadership vendors. However, Symantec's calculation is obviously empty, because Sun has decided to acquire software companies specializing in developing Java compiler technology, and preparing to develop their own JIT compiler, which is later Sun Hotspot compiler technology. In addition, Microsoft and IBM have also begun to joined the Java Jit compiler's race. In order to compete with Sun to compete with the Java leader, IBM not only develop IBM JDK, but even develop IBM's Java Jit compiler. Strictly speaking, IBM's Java JIT compiler quality is much better than Sun, not only stable and executed is much faster than Sun, so that Sun is also light, it is not a taste. It can even be said that IBM's Java Jit compiler quality is not from where the Java Jit compiler of Symantec. What is more trouble is Microsoft to make IE to compete with Netscape, and can also perform Applets, so it also starts to develop a well-developed Java Jit compiler. Especially after Microsoft gets Anders Hejlsberg, there has been a major breakthrough in compiler technology. Although Microsoft's JIT compiler has been unlike other manufacturers' Java Jit compilers, their quality is quite excellent. In Microsoft constantly improved, according to my test, the speed of Java Bytecode after its compiled Java Bytecode is the fastest, even the JIT compiler of IBM and Symantec is not an opponent. So from my point of view, in this Java Jit compiler, Microsoft should be a champion. If Microsoft does not continue to support the latest JDK standard, it is mixed some Microsoft's own things, and it is likely that the most widely used Java Jit compiler is the Microsoft's JIT compiler. As for Symantec, after obtaining the advantages on the surface of the JIT compiler, the focus immediately placed the development of Java Bytecode into native applications directly into native applications. Saraday Symantec successfully developed this compiler, so Borland is very nervous and ready to follow up. Symantec also adds this native Java compiler into Visual Cafe to become an attractive function. However, this feature has caused many Java users to criticize, because they think this violates the spirit of Java "Write Once, Run Everywhere", so that manufacturers must develop native Java compilers for every different platform, This will cause inconsistencies of the Java application to perform in different platforms, and the application of C / C language developed in different platforms is different from the same problem. Later, even Sun did not agree with this approach. Of course, this is because Sun wants to push your own HotSpot compiler technology. Therefore, the native Java compiler is no longer attractive after a short period of time, and Borland is originally awarded the Java compiler of JBuilder. Threats of Microsoft VJ In 1996, Anders Hejlsberg came to Microsoft's first work soon, that is, Microsoft VJ . VJ is about to be launched, a big shock for many software companies. For Sun, this is the challenge of Microsoft in Java.
Under the dilemma of Sun's Java development tool, it has to face Microsoft, which is good at developing tools, especially the boutique developed by Anders. For other Java development tools manufacturers, it is also a hanging bile. Visual Café After JBuilder, Visual Age for Java launched, the market share has been slowly saved, and now you have to face Microsoft's competition again, and the shadow of Symantec C / C has wrapped his heart. And Microsoft's death IBM is even after Visual, Visualage for Basic, it is difficult to launch Visualage for Java, preparing to play a good game in the Java Development Tool market, and did not think that Microsoft now spoken. For Borland, this news is even more disturbing, because Borland itself's Java development tool is still in the R & D stage, yet launched, and it will be the last Java development tool in the market, and the main competitors have been Many. Now Microsoft actually introduces Java development tools earlier, and is developed by Anders Hejlsberg. Borland, of course, know the strength of Anders Hejlsberg, naturally not daring VJ influence. What is even more troublesome is before VJ , Microsoft has always kept vague attitude towards VJ , and is reluctant to indicate whether VJ is a pure Java development tool. More than Borland is surprised that Borland's tests for VJ beta indicate that the program code compiled by VJ is actually implemented quickly than the native Windows development tools such as Delphi. This means that VJ may have a serious impact on Java development tools, but even threats to general Windows development tools. However, Borland Analysis If VJ really begins to threaten the Windows development tool, the VB will be affected by the most developed tools. But Borland is still worried because VJ may still have a certain impact on Delphi and C Builder, which is Borland unmanned. Of course, this also accelerates the determination of Borland to develop Latté because they can no longer be dragged. I remember that I was also talking to Borland in the Asian Singapore R & D headquarters Mr. Inn Nam Yong's performance and anxiety for VJ may have an impact. Mr. Yong also said that VJ performance made them surprised. It seems that Anders Hejlsberg has undergone hardships in VJ compiler technology, and its performance has long been more than the general Java compiler technology, which is indeed a scratch and is more troublesome. It can see Delphi's figure from VJ . Borland's R & D has already understood this situation, Borland's compiler team is also researching related issues. It can be seen that Borland has already prepared related technologies at the time, and it has mastered the initial situation. Microsoft VJ is finally officially launched in November 1996. The whole world is also waiting. It is ready to look at VJ will generate how much destruction power, and Sun is more ready to see if Microsoft will violate the Java protocol between any Sun and Microsoft. . Of course, Sun is worried about Microsoft to destroy Java development.
VJ has got some essays at the beginning, after all, this is Microsoft's Java tool, and software personnel who use Microsoft Development Tools will of course consider VJ . At the same time VJ also attracted some developers who want to use Java language, but still plan to stay in the Windows platform. However, VJ has also been quickly subjected to all Java development tools and a comprehensive encirclement that supports Java platform vendors. They are afraid that Microsoft's invasion of the Java market will make other manufacturers unable to survive. After the Sun also begins to lead Microsoft because Sun is afraid of Microsoft to slowly dominate the Java platform and standard, and Microsoft is very technically gradually destroying Java language and standards. For example, VJ provides many non-standard. Java usage and clearly tied VJ to the Windows platform, destroying the dream of Java's "Write Once, Run Everywhere". Moreover, if Java developers use VJ , then they will be inseparable from the Windows platform. The Microsoft plans to limit the free choice of the developer by providing first-class "class Java development tools" to limit the free choice of the developer. Due to Sun's lead criticism, you want to use Java's developers and companies soon find VJ is not a standard Java development tool, so the enthusiasm for VJ will soon retreat. VJ threats to Java and Windows development tools are also released quickly. VJ is probably a DOS version of Microsoft Pascal after the second time in the development tool after the DOS version. However, in my point of view, VJ is essentially an excellent product, whether it is compiler, Framework and integrated development environment. The reason why VJ is defeated is really because the situation is strong, the Java platform is also the first time that is not the market dominated by Microsoft. Under the joint attack of the Java coalition, even the software giant also avoids tripartite. Because the first time I was hit on the Java, I would like to make Microsoft in the Java platform to become climate, making Microsoft to fight on Sun, and fully fired in the Java market, thereby causing Sun accuses Microsoft violating the SUN The results of the Java contract, while Microsoft simply simply removing Java from the operating system. Of course, this is the battle between the Java platform between Microsoft and Sun, which has exceeded the scope of this book, and perhaps the entire process should be described by Microsoft or SUN. Although afterwards prove VJ in Java development tools fail, Anders Hejlsberg spent the heart of VJ , because VJ compiler technology and Framework and integrated development environments are incorporated into the Microsoft .NET program In the core technology. For example, C # language and Java are very similar, C # compiler technology wants to also borrow many of the original VJ excellent compiler, so the optimization of the C # compiler is also better than some of the original Windows development tools. Compiler level. Anders Hejlsberg's efforts are activated in Java and .NET's frontal battle. IBM Visualage for Java's launch of IBM's performance in the PC Development Tool market has always shook, because its style of "playing" is always unable to use its development tools.
However, maybe the IBM's signboard is too big, plus it will send IBM development tools free to buy IBM machines or software, so someone always uses IBM development tools. After I was less than IBM Visualage for C / C , I went to thank IBM's development tools. IBM must not give up Java's potential market, because for IBM, Java is not only language and development tools; more importantly, the Java platform involves the competition between IBM and Sun in huge business opportunities and big customers. . IBM is not just to support Java, but also want to get Java's main control from Sun Hand, so IBM is not absent from the important Java development tool market. IBM has quickly adopted many of the very popular elements in Visualage for C / C as the basis for developing Visualage for Java. For example, Visualage for C / C project management functions, component designers, etc. In fact, the readers who use Visualage for C / C will find that Visualage for Java has a very intimate sense. Not only do all buttons are circular shape, even when the feeling is slow, the performance of the entire integrated development environment is very similar. Due to the partial concept and program code of Visualage for C / C , add IBM has the most abundant resources, so Visualage for Java has developed very quickly. In September 1997, IBM finally launched Visualage For Java, started directly with Sun, Symantec. After IBM launched Visualage for Java, Borland is destined to become the last manufacturer launched a heavyweight Java development tool. However, IBM's competitive target is obviously not a manufacturer of Symantec and Borland, but SUN and Microsoft. IBM has taken several parallel strategies in Java technology, hoping to get the lead position in the Java world, as this is related to the IBM's largest business - hardware sales, service providing and IBM operating system sales. If IBM can make decisive status in the Java world, you can erode the market of Sun, and the most unjust situations do not want our customers to naturally think of Sun. As for another locking target Microsoft, IBM is intended to hit or suppress through Java increasing moment. Therefore, IBM is signed by the SUN to sign the Java contract, and the legal authorization of Java is obtained. On the other hand, it is also invested in a large number of R & D resources to develop its own JDK version to facilitate porting to other IBM's exclusive platform, and do more than Sun's JDK. Stable and efficient, then let Sun and IBM have always disputed. Then IBM introduced Java development tools, again and Sun's Java Workshop competition. However, in terms of characteristics, Visualage for Java-locked customer base should be IBM customers, large business customers, and other customers who are directly competitors, such as customers, and HP customers. Visualage for Java requires a relatively strong machine to perform, and the start version is very focused on the support of the team development, unlike other Java development tools, it is paying attention to convenient, practical features at the beginning, and gradually strengthens the team development later. These differences are IBM wants to compete for prove that larger customers. This is also available from later many professional media and magazines to evaluate the Java development tools, Visualage for Java has got the highest evaluation in team development functions. Due to the unlocked customer base and Visual Café and JBuilder locked customer base, there is no serious competitive conflict at the beginning of Java development tools.
However, with Visual Café and JBuilder gradually struggled to enter the corporate market, and Visualage for Java started to reduce the market in order to expand the market, it would not be avoided after the next Java developing tools. Section 2 - Java Integrated Development Environment War Borland has gradually been mastered after the development of Java Jit compilers, has gradually been mastered for Java technology. After Java Workshop, VJ and Visualage For Java have been launched, Borland knows no longer delay JBuilder's launch time, otherwise it is destined to exit the market of Java development tools. Therefore, Borland has finally launched the last passage of JBuilder's R & D team, be sure to launch JBuilder in 1997. After the first phase of the competition, the JBuilder team mastered Java's JIT technology, and immediately broke the development of JBUILDER. Although Java is a new language and revolutionary platform, the development tools are not even more than an interpreter, integrated development environment, data inventory ability, Framework, and other tools and PLUG-IN. At that time, there were many members who were transferred from previous Borland C / C , and some members included Borland original software research members, Paradox members, Visual DBase members, and new engineers found outside Borland. When Borland develops JBuilder, since Java has no complete component architecture, there is no data perceived component standard to easily develop Java database applications, there is no complete Java visual component, so Borland decides to develop a set of component groups first. Let JBuilder have the best component development capabilities. This is a technology that Borland is good at, because Borland wants to develop a set of Java Framework for Java, which is the origin of JBCL (JavaBeans Component Library, and JBCL's architecture will also become the basic technology of JavaBean in SUN. Architect, which was responsible for the JBCL architecture was Mr. Joe Nunoll. The handsome guy was originally located in the Paradox team, and after Borland gradually lost to the desk-type database battlefield, it turned to the Latté team to design the Latté components architecture. JBCL's main implementation engineers were design and realizing the General Engineer Carl QUINN of Borland C / C Framework-OWL. Carl Quin has rich experience in component design and Framework, and OWL is also a boutique in terms of technology. Therefore, after the Borland C / C product line stops, Carl Quinn is converted to the Java runway by C / C , after all, C / C and Java are very similar. Carl has a wealth of experience, and it is flexible to lead the development of Latté's Component Framework. Due to the efforts and results of Carl in JBCL, it is responsible for the development of Borland's Java component model BAJA. Later, Carl was naturally invited to Javabean's development group after Sun's core basic technology using Baja as JavaBean, using Baja. Since Borland excellent performance in Java components, SUN and BORLAN have also opened a close cooperation. Although Borland puts a time in Java, Borland is slightly tight, but it is slowly on your head with a solid technology. Latté's Framework developed a stable development under Joe Nunoll and Carl Quinn. In fact, the performance of JBCL has always been very excellent. When Latté is subsequently launched, JBCL is also one of the important functions that let Latté stand out.
Joe Nunoll and Carl Quinn's credit can be said that it is not small, and I admired Carl Quinn once again proved his solid technology and experienced experience in Framework in Java. Although Java Framework is important, it is only one of the struts of the entire full development tool. Latté To launch features that still need compilers and integrated development environments. In the Java compiler, Borland and end entrusting Dr. Niklaus Worth research team developed the Java Jit compiler, the Latté Development team began to work in research. At that time, the JIT compiler has been fully on, and Borland has not been launched in Latté. If there is not enough resources, it is not only the winning of JIT, but it may seriously affect the launch time of Latté. Therefore, the Latté team decided to first develop a Java compiler that compiles its compilation, and can be matched with the Latté, and no longer seek the efficiency first. From now on, the decision of the Latté team was very correct at the time. Because: First, Borland did not have too much bullets; second, Latté's time course no longer delay; third, the most important thing is that SUN announces the development of the HOTSPOT compiler technology, suddenly All the winds of all Java Jit compilers are taken away by Sun. Especially after SUN decided to build HotSpot in JDK, competing for the Java Jit compiler not only meaningless, but also has no added value for Java development tools. Therefore, after Sun's HotSpot compilation technology revealed, Symantec quickly tired in the Java Jit compiler market. After the Latté team determined the Java compiler's strategy, he immediately took over the follow-up development of the Dr. Niklaus Worth Research Group, and developed the Latté compiler, Latté's PLUG-I in the integrated development environment of the Java compiler to the Latté. And also performing research work for Java compiler optimization. At that time, this Java Compiler Development Team led by Carl Fravel and others, and also included Borland's compiler group. Carl Fravel is primarily responsible for the Latté compiler and the compiler's PLUG-IN software, and he also participated in the development of the Latté database. At that time, Java was still quite vulnerable in data inventory, and Borland decided to first borrow the Delphi's performance in the database, which enables Latté to install the Delphi's various database drivers through JDBC standard interface. Connect to the most database, this is later jbuilder's DatagayTeway. In addition to Carl (um, the Latté Development Team has two Carls), Sergio Cardoso is also a member of the Latté compiler. Sergio was originally one of Borland C / C Ten developers, specializing in C / C optimized technologies, and transferred to the Java product line like Carl Quinn et al. Sergio and Carl Fravel work together to build the core engine of Latté. After latté listed, in fact, every version of JBuilder compiler has progress. As soon as it is, compiling the Java application is quite fast in JBuilder 7/8. This shows that JBuilder is the same as the original Borland C / C , which is quickly enhanced in the initial version, and continuously improves the speed. At that time, Latté did not have a true total architect, only ARCHITECT in different technical fields, such as Joe Nunoll. Therefore, the main helmsman at that time should be a product manager and a senior researcher of Borland's Java.
The product manager and the Senior researcher of Java will discuss the development direction of Latté in all the fields of ARCHITECT. Therefore, in the initial development phase of Latté, its mode is like a collegial system in the Roman time. The Latte product manager at that time was Klaus Krull (k. K.). This dear brother grows high and big, and it is also a quite enthusiastic person. K. K. Originally a member of the Paradox and DBase group, responsible for the product line of Paradox / DBase. After the Latté team was established, K. K. immediately jumped to the Java product line. In fact, many PARADOX / DBASE engineers also want to be converted to Java products, so there is a lot of wind waves inside Borland, which has also caused many people to leave. This story will be described in detail in a later DBASE-related chapter. K. K. After joining the Latté group, he actively led the Latté team forward. Perhaps because K.K. is not good in Paradox and DBase, so I want to prove my strength by Latté. However, K. K. people may be very good, but management seems to be still not very spirited, and the initial progress in Latté is still slightly slow. After IBM Visualage for Java was launched in 1997, the Borland high-level order K. K. Never delayed progress, it must also launch the first Latté version in 1997. Fortunately, Delphi 3 won the winner, while K. K. is a good buddy of Ben Riga at the Delphi product manager, so the resources of Delphi transfer to Latté will end, let Latté slowly catch up. As for the main figure of the Latté architecture, the BLAKE Stone is not aware of everyone, because Blake Stone is added to Borland after latté. The initial architecture of Latté should be Steve Shaughnessy. Steve is a senior researcher in Borland, and it was also one of the people who were first invested by Borland into the field of Java technology. However, one of the shortcomings of senior researchers is to constantly imagine and study software technology, but the mastery of product progress is not their expertise, nor their most concerned things. This is why Latté started with a very slow development schedule. Until the last K. K. joined and faced all the resources and time wanted to catch up with all resources and time after Borland high-level and market huge stress. When the Latté team develops JBuilder's first version, you want to learn the use of Delphi's successful Open Tools API feature, which is a complete and very flexible open integrated development environment for JBuilder. However, since Delphi's Open Tools API still does not have a large number of architectures that use interface programming, considering Java has a well-defined interface mechanism, it cannot fully use Delphi's Open Tools API design, so starting JBuilder integrated development environment Add- INS function is developing very slow. At that time, the engineer of the Add-INS function in the JBuilder integrated development environment was outside Carl Fravel, and another major engineer was GREG COLE. When Carl Fravel and Greg Cole understand that after using the Delphi Open Tools API to design JBuilder's Add-INS architecture, it will begin to develop JBuilder itself integrated development environment open architecture, and directly use the mechanism for interface programming to design JBuilder. Open architecture, this is a different place in Delphi at the time. After the Danny Thorpe has joined the interface mechanism to the Object Pascal program language, Delphi's Open Tools API launches the 2nd wave of bigification, using an interface mechanism to redesign, which is later Delphi's famous OTA architecture (Open Tools) Architecture).
In November 1997, Latté finally completed and launched the market. The formal product name is set to Open JBuilder, which is to emphasize that Borland's Java development tool is just like Java itself is an open architecture. After the Open JBuilder 1.0 is launched, the Java Development Tool market is finally competitors gathered together, and each manufacturer has finally made the real market leading market leading market faucet. Shortly after Open JBuilder 1.0, almost all information media and Java's professional magazines have javas java development tools, and want to compare all Java development tools, and let Java's users understand the boss of the market. Visual Is Café to face the challenge of emerging forces, maintain the status of the market. At that time, most magazine evalued the target included Symantec's Visual Café, Sun's Java Workshop, IBM's Visualage for Java, and Borland's Open JBuilder 1.0. For Symantec, the first Java Development Tool General Assembly war is in the situation of Yiwu, and Visual Café is also the only complete use of C / C language in 4 Java development tools, so the Java compiler is not good enough. The Quality of JVM does not have a very obvious advantage of the implementation speed of Visual Café, of course, and other competitors. The only disadvantage of Symantec's Visual Café is that the Java program code is added to the developer code, which causes the Java program code written by Visual Café to not use the consequences of other development tools. And Visual Café still has unusually accurate problems when the Render Java graphic user interface. The evaluation of Sun's Java Workshop was relatively conservative. After all, Java Workshop is the product launched by Java Authentic Manufacturer Sun. Although it is better than competitors in terms of function and execution efficiency, and a small problem is a lot, but in order to give Sun's face, the media still gives too much demographic. There are even some media, also praised Sun with courage to develop a Java development tool that completely uses Java language to prove that Java to develop large applications. However, although the media and magazine gives a sun-face, Java Workshop fled the market's test, and slowly withdraws the market of Java development tools. Although IBM's Visual For Java at the time, although the most slow Java development tool, the performance in high-order function is far from all competitors. Visualage For Java team development feature, project management function, and visual designers have greatly surpassing other Java development tools. However, Visualage for Java uses the exclusive format, so its program code is not easy to use in other tools, and the Visualage For Java project has also happened through the REPOSTORY, and therefore, Visualage for Java is easy The score in use is more than other competitors. For Borland's Open JBuilder 1.0, the latest outcome of this tight to enter the competitive market is unreasonable in the first collective ratio. Original Java users and professional media have a high expectation for Borland's products. Since Borland has always been refined with development tools, it is concluded that most people think that Open JBuilder should be the most sufficient, but the results after the evaluation are not. First, Open JBuilder is not a development tool that is purely written in Java, but mixed Java and Delphi program code.
However, the final implementation efficiency is not only more than Visual Café, but is not much faster than pure Java Workshop. In addition, Open JBuilder is not a Visual Café in terms of Visual Café, which is not the opponent of Visualage for Java in terms of visual designers and high-order functions. Open JBuilder made the big mistakes at the time, and of course I also included me. Therefore, in most comments, Open JBuilder only received medium evaluation. Of course, this result is also reflected in the market performance of Open JBuilder. Is HOTSPOT compilation technology a joke? In 1997, the market gradually became more and more Java development tools, and more and more people began to try to use Java, but they also had more and more people complaining about Java's execution efficiency. At that time, the PC did not have a 1GHz execution efficiency and 512MB RAM memory, which was a painful thing to perform Java at the time of the machine. I still remember that I had no power enough machine to run Open JBuilder. Every time Open JBuilder, I can't stand it. At that time, I also joked that the machine from executing Open JBuilder's integrated development environment in Open JBuilder, I have long enough to use Delphi to write a program. The main reason for the slow efficiency of Java is of course the Quality of the Java compiler and JVM is not excellent enough. In order to rush to let the information receiving Java a standard, Sun must find ways to overcome this issue. While overcoming that Java performs slowly, almost all things that support Java software vendors want to solve, but Java's authentic manufacturers Sun are responsible. It is also because Java performs slow efficiency, and many small software vendors have developed various technologies and compilers to improve or solve this fatal disadvantage of Java. Soon Sun found a small software company, this company developed "Adaptive Compiling" technology to speed up JVM execution efficiency, and use similar techniques to improve the quality of the Java compiler. SUN immediately decided to purchase this company immediately after learning these outstanding technologies, and achieve Sun's next-generation Java compiler and JVM according to their technology, which is the origin of Sun HotSpot technology. SUN has been put into new Java compile technology, there is a preliminary result. According to this new technology, Java Bytecode and new JVM execution efficiency have improved much more than before. This makes Sun more confident, and immediately announced this new technology to the world and named HotSpot. Sun claims that the last Java compiler and JVM will provide an effective efficiency similar to C . After Sun announced HotSpot technology, it immediately caused the world's Java user's fanatics. People think that once Sun launches this technology, Java will be expected to overcome the last shortcomings and thus a uniform world. At the same time, this also caused very large discussions and controversies in the information industry. In particular, the C / C community believes that this is impossible, although "Adaptive Compiling" is very creative, but it is impossible to compare with the C optimization compiler that has already existed, Java's Bytecode is impossible. Beyond C . But from SUN's some HotSpot compilation numbers announced in it, "Adaptive Compiling" is very promising because it improves the amplitude. So the world-related personnel are preparations for the last results of the Sun. After Sun's first announcement of HotSpot, let all Java users are looking forward to, hate SUN immediately launched this technology, and disseminate the pain of Java.
However, over time is close to, Sun is announced in the last level and announced that the time course launched by HOTSPOT is delayed. After the work of software research and development, he didn't blame in the software world. At that time, there was no too much dispute, but not only let Sun strive to have more time, but also successfully launched the launched time. However, Sun is still unable to launch HotSpot Technology when Sun announced the second time course expired. Soon Sun has to declare the time to delay Hotspot. In the case where Sun continues to jump tickets, I finally started to laugh, and I said that Hotspot is simply a scam, and Sun is simply unable to introduce Java compilation technology close to C execution efficiency. In 1999, Sun self-knowledge will never push the time of Hotspot launched. So in August, the leader of the HOTSPOT R & D team (a doctor, but I have forgotten his name) In Borcon, I will introduce HotSpot Technology to the people participating in Borcon and showing Hotspot on site. R & D results. Although everything looks great, when the audience of the scene directly asks if the HotSpot technology can exceed the efficiency of C execution, this doctor does not answer, only explains that HotSpot is indeed perbated in some applications. effectiveness. After I listened, I already know the final result of HotSpot. Since HotSpot is forced to launch the market, everyone quickly learned that HOTSPOT and C execution efficiency is still a distance, and it is not possible to exceed C performance. This caused some of the newly expected C / C programmers to return to the C / C language, and did not switch to the Java market. This is why the C / C market has been affected by Java, but there is still a lot of users and markets, and there will be a lot of C / C programmers to enter the Java market like many people predicted at the time. Because Java cannot completely replace C / C language to complete some work. And what? In order to transfer everyone's disappointment of HotSpot, the R & D focused to the development of R & D to the Internet / Intranet, EJB component model, and Java Mobile system. The hotspot boom in the sensation is also gradually falling. Now SUN does not mention the HotSpot compiler, just continuously improve the compilation quality of HotSpot in each new version of JDK. I remembered that Sun is the most sigh of Hotspot or the world. However, HotSpot is not a place, which is indeed refined a lot of Java Bytecode's quality and JVM execution efficiency, but did not reach the extent of Sun boast approaching or transcend C language compiler. In the current situation, Hotspot enables Java's compilation quality in the servo-based efficiency, providing very good execution efficiency. However, when the client, especially in the application of the graphical user interface render, it is still quite slow. For the case of Borland itself, Borland uses Javoker for Java's execution efficiency that use Javoker for Java, is quite close to the performance efficiency of Visibroker for C / C . So if you use a good JVM with good quality, then according to the test data inside Borland, Visibroker For Java exceeds Visibroker for C / C in some specific applications. HOTSPOT is not a "joke technology" who hangs so many years after so many years? Different people may still have different answers now.
Borland's dilemma and options Open JBuilder have been launched in 1997, but the reflection in the market is not as good as expected. Of course, this is a lot of reasons. The first is Open JBuilder launched too late. The initial Java market has been occupied by other Java development tools, especially Visual Café; the second is that Open JBuilder has been invested in the market, so there is nothing special when competing with other Java development tools. The prominent function, obvious advantage, is certainly less competitive; the third is that Open JBuilder uses Delphi and Java program code at first, so Open JBuilder activation and the form of the form must be slow, unlike Visual Café The Java development tool written in pure C / C program is rapidly, thus causing a bad impression to many programmers. Although IBM's Visualage for Java is also very slow, it supports very good in high-level team development, and usually uses team development functions to be large, large users, so the machine is also very useful. Ok, it is also acceptable for the slow response of VisuaLage for Java. Open JBuilder is not as expected, which makes Borland urgently because it cannot withstand the loss of the Loss of the Java Development Tool market. Therefore, there are some discussions in the Borland's Java Development Tool R & D team, that is how to make Open JBuilder can last, and the fruit of victory. For the failure of Open JBuilder, the developer of Open JBuilder began to reflect on whether Delphi used Delphi as Visual Cafe, so that the Open JBuilder's execution reaction accelerated to the user acceptable, because the borland can't be Accelerate the execution speed of Java. Also using Delphi Development Open JBuilder's form designers can also avoid many JDK's bugs, which will not affect the development cycle of Open JBuilder due to Sun development or improve JDK. This idea has attracted great controversies within Open JBuilder. The integrated development environment for rewriting Open JBuilder using Delphi can have many short-term benefits and products will have obvious improvements, you can have the capital of other competitors. However, people who oppose believe that the use of native development tools to develop Java tools is to walk back. These people think that Java will be able to develop to a mature stage, and Open JBuilder will have the final victory. Now it is only a time setback, there is no need to be discouraged. For Borland, how to continue Open JBuilder is a difficult choice, because Borland urgently needs income, and Open JBuilder's research and development cost is amazing, and it is very hard to support Delphi. However, if you return to use Delphi development, you may lose future opportunities, how should this be decided? Java genius joins all Answers gradually gradually clear after the new product architecture of Open JBuilder. The Blake Stone was originally the technical supervisor of DSW Systems Corporation, while DSW has always interacted with Borland, and many DSW people have been responsible for technical lectures in Borland's Conference (Borcon). Mr. Blake Stone is also responsible for a lecture in Borcon in 1997.
