BORLAND C / C counterattack
When Microsoft Visual C 1.0 wakes up in the C / C development tool market, Borland wakes up from the victory dream of Borland C / C 3.1, thinking about how to face Visual C violent power. In fact, Borland, if the head is awake, look at the market of C / C development tools, then Borland should find that although Visual C has been more than 2 years, the strength has been large. However, Borland C / C 3.1 is still in many ways and Visual C . For example, the optimal compiler of Visual C is still behind Borland C / C 3.1, and the second point is that the MFC still has no complete packaged Window API, and the MFC is encapsulated in a lower order, is not very object. Guide, not very easy to use. In fact, in my point of view, I think it is because MFC is not easy, so Visual C needs to provide the function of the MFC program code in the integrated development environment. The third is that Visual C is not very good. The Container Class of the package information, while Borland C / C has a very easy to use BIDS category library. The fourth, and most important, Borland C / C 3.1 still has an extra large market, and almost all peripheral public programs, Shareware, etc. are developed using Borland C / C 3.1. So if Borland is not anxious, develop next-generation C / C development tools, even if Microsoft Visual C can plunder some market share, but if the next generation Borland C / C can immediately pull away like Borland C / C 3.0 The distance of Visual C / C , then Borland will still have the status of the king in the C / C market.
It is a pity that maybe Philippe Kahn is scared in a business with Microsoft's FoxPro for Window, so it is eager to immediately introduce new Borland C / C to pull back the face after Visual C / C 1.0. However, Philippe Kahn forgot that Borland lost many people during this time, Eugene Wang also left, but more importantly, in the past three years, Borland has almost no continuous development of the next generation Borland C. / C , how can we be able to launch a product in a short time?
However, Philippe Kahn can't manage so much, and hurriedly found a Carl Quinn and others to immediately develop the next generation of Borland C / C , so Borland C / C 4.0 is in such a duck down, rushing down. Borland made a lot of tricks when developing Borland C / C 4.0. First, Borland determined the new development and integration development environment in such a short period of time. The second is to completely rewrite OWL. The third is a sharp modification of the optimization compiler. The fourth is to integrate the tricky VBX, Borland actually let 16 yuan And 32-bit prices can use 16-bit, ugly VBX simultaneously.
Every item mentioned above is a big project, Borland should start doing these work after Borland C / C 3.1, now you have to revisit such a complex C / C development in just more than a short year. Tools, almost impossible work. But under Philippe Kahn, these Borland's engineers have made a hard scalp. But I have to be very sorry, when I tested Borland C / C 4.0 in Beta to say that if Borland rushes launched Borland C / C 4.0, then not only produced any impact on Visual C , Instead, it is the behavior of suicide, because the bug is too much, the reaction of the entire integrated development environment is also very slow, its optimization compiler is more joke, and it is a mistake. I also joked, is it because Microsoft digs a lot of Borland C / C talent from Borland, so winning Philippe Kahn is also in color, from Microsoft anti-Microsoft C, but unfortunately digs to Microsoft C 4.0 People.
But it is clear that Borland does not hear me, or other Beta testers, more than 1 year after Visual C 1.0, 4 years after Borland C / C 3.1, Borland finally launched a new generation of Borland C. / 4.0, this shoulder and Visual C 1.0 confrontation C / C development tools.
On the occasion of Borland C / C 4.0, Borland did have a great creation of 4.0, I remember that in all important computer magazines at that time, such as Byte, PC Magazine, Dr. Bob, etc., there were 4.0 full page advertisements. . The content of this advertisement is based on a huge owl. It is the mainland of Borland C / C 4.0, which is mainly because of OWL, but I can't find the advertisement now. .
Borland C / C 4.0 in Jiangshan
At that time, Borland used the following advertising words:
"Visual Is Only A Facial Facade"
To satirize Visual C / C , only the basic wizards of the MFC program code, while Borland also provides similar features to provide similar features to generate the OWL program code selected by the user, Borland C / C 4.0 The integrated development environment also provides a 3-side version of the visualization that allows you to fully master the entire project.
For example, in the figure below, it is an eye-catching APPEXPERT:
Remember the APPEXPERT provided by Borland?
The following picture is a registered trademark of Borland C / C , a 3-side version of the window development environment. Seeing the picture below, I remembered the days to use C / C writer, the following code code version clearly shows the smart Window schedule system written in Dingxin in Dingxin in 1995, time is really fast. what.
Missing Borland C / C 4.0 integrated development environment, three-sided window
At that time, Borland C / C 4.0's 3-sided integrated development environment was created a new situation because this integrated development environment allowed programmakers to know the window message defined by each application, and can immediately display the program window below It is indeed very convenient, and it is more advanced than the integrated development environment of Visual C / C at that time. Add Borland's more advanced compiler technology and architecture better C / C Framework-OWL. The character says that Borland C / C 4.0 should get great victory, then why will it finally take a failure? That's true, the order is really like a snow film after Borland C / C 4.0, the sales situation is very good, because this is Borland's masterpiece after several years, many Borland users can't wait to upgrade, just like At the beginning, I also desperately demand Taiwan Borland to give me Borland C / C 4.0. However, after a period of Borland C / C / C / C / C / C / C / C / C / C / C / C / C / C / C / C / C / C / C / C, the market will be quickly cooled, because various negative criticisms are emerging, this is the main reason is because Borland C / C 4.0 is really bad, just like front. I said in the beta test, because Borland is too eager to launch 4.0, there is no many errors in the final stage, and there is no work of fine-tuning, and there is too boldly to join too many advanced technology, causing the entire product. Unstable, but caused a big mistake. The following should be the main reason for the original Borland C / C 4.0 Waterloo:
* Integrated development environment - there are too many bugs, easy to get rid of slowly * Compiler - optimization to play fire, generate errors, compiler code * OWL - use a new multi-inheritance architecture, although it is correct The practice, but the OWL in Borland C / C 3.1 is incompatible, causing many programs that cannot upgrade C / C project * VBX - bold use of technologies that can use VBX at 16/32 yuan, causing some VBXs that cannot be Working in Borland C / C 4.0
I think it's a pity that OWL is OWL, because OWL 2.0 has a first-class performance in all aspects, it is really a strong competitor of MFC, OWL 2.0 has also obtained the uniform and praise of all parties. Helpless because OWL 2.0 has changed from the basic architecture, this is to solve the problem of the original OWL 1.x used an inexpensive C / C compiler technology, but this has caused the original Borland C / C master. It is troubled because the upgrade is not easy. For new C / C users, because Borland C / C 4.0 itself is unstable factor, it has caused OWL 2.0 called the next game, it is a pity that the OWL team is working hard.
I remember that my discounted spreadsheet VBX component used in the past, because it has been useless in Borland C / C 4.0, and will cause the application to be removed, and finally tracking the executive code is found to be Borland C / C . 4.0 The problem, so finally, in the curse, the use of 4.0, and returned to Borland C / C 3.1. I thought that the people who used Borland products were unable to endure the quality of 4.0, how can other programs use this product. I think this is why it is the reason why 4.0 is completely defeated, because Borland has launched a product that is fundamentally unbearable. In my time in Borland, I have a Vice President of Singapore and now the vice president of the Borland Developer Relationship Department. David is also very emotional, David, "We Screwed IT Up!", " IT's a mess. David and said that the entire Borland C / C development team is very confusing, and the development team in the past Borland C / C 3.0 / 3.1 is much difference, except for some important characters to leave Borland, and Microsoft also digs Outside of a major ticket, Philippe Kahn's direct intervention, causing people to have a lot of reasons.
David I. Say "We Screwed IT Up!", "IT's A Mess"
After Borland C / C 4.0 quickly lost, Borland also considered the seriousness of the problem, so immediately developed Borland C / 4.0 Patch, which was called Service Pack. However, there is no complete solving problem in a later version 4.01, and it has been to solve some serious problems with 4.02. It is not allowed to give, the time is too long, the market has gone a huge change.
After Borland C / C 4.0 lost, serious consequences immediately, first of all, the Borland C / C market is large and fast, allowing Visual C / C to grow fast. The second point is that Jiangshan, which was labeled in the utility market, was also in the public program market, and many hardware vendors also used the Borland C / C 3.0 / 3.1 written driver and started to Visual C / C , while serious In the application market, due to the quality of 4.0 and the relationship of OLE, it has also begun to use Visual C / C to write an application.
Borland has been defeated in three main application markets, and the Jiangshan, C / C is destined to be the owner, and the trend is unparalleled.
Borland C / C , Visual C / C , Watcom C / C and Symantec C / C
Since the defeat of Borland C / C 4.0, Borland's huge fortress in the C / C market seems to be no longer downtown. Visual C / C is constantly receiving the Borland C / C lost market. At this point, two strong opponents have also been added in the C / C market, which is Symantec C / C and Watcom C / C .
Development history of Symantec C / C
It is also one of the two opponents to say that these two opponents are not small, first say Symantec C / C . Its Think C / C is very famous compiler in Macintosh, so there is a deep foundation in the C / C field. After Symantec mergers, the first C / C compiler ZORTECH C / C , Symantec entered the PC development tool market was also on the strings, but it was unfortunately that Symantec has not found a rich experience in the PC. Tool leader.
Perhaps God is destined to cause a later C / C compiler! At this time, Borland C / C 3.1 behind-the-scenes pillar eugene wang just turned over and left Borland. Symantec is not lost at this time, immediately introduces Eugene Wang to Symantec, and introduces the first C / C development tool for Symantec. In 1993, Symantec C / C launched the first Symantec C / C version under Eugene Wang's helm, and immediately won praise from the market. Since this, Symantec C / C military is very large, continuous continuous improvement, and gradually gains a small C / C market, which is thought to fight against Borland C / C , Visual C / C . At that time, Symantec C / C was highly recognized by the most gorgeous, advanced integrated development environment, and the performance of the C / C compiler is not lost to other compilers.
At that time, I wrote a C / C article on the run! PC, so Symantec C / C also had a connected with me, and gave me a set of up-port Symantec C / C , I hope I have written C / C articles for Borland. In addition, you can also write something for Symantec C / C . I think this is one of the benefits of writing a technical article, that is, you can get many of the most HOT development tools. I still remember that after installing Symantec C / C , it is indeed attracted by its integrated development environment, because it is great, Borland C / C and Visual C / C integrated development environment and Symantec C The integrated development environment of / c has become a tasteless, ordinary, and now I still have the thumbs to say "Zhu." "Zhu." "Like" the integration of Symantec C / C . I want Eugene Wang to build Symantec C / C in such a short period of time, in addition to prove his extraordinary skill, there is also the meaning of philippe kahn. Prove that Philippe Kahn made him leave Borland as a wrong decision. The reason why I am because of the object of Symantec C / C like the noddist challenge is Borland C / C .
For me, Symantec C / C is like a skillful spirit, and equipped with a gorgeous C / C army.
Development history of Watcom C / C
It is very interesting that Watcom C / C walks and Symantec C / C are almost completely opposite. At that time, the WATCOM C / C compiler was a small company in Canada, but this company has an in-depth study of optimal compilers. At that time, Watcom C / C was known to be famous in DOS, and many manufacturers who wrote games and DOS EXTENDER were named Watcom C / C at the time. Because BORLAND C / C or the best chemical code generated by Visual C / C is more than one case in the optimal chemical code of Watcom C / C . Then, the most famous DOS Extender manufacturer Pharlap was also using Watcom C / C , so Watcom C / C was a special C / C development tool in the professional C / C programmaster and system program. I don't know how many people remember this company, or have anyone who remember Andrew Schulman's great software technician. At that time, Andout Schulman's undocument Windows book was red over half a day, and Microsoft wants to tell Andrew Schulman. Andrew Schulman is the chief engineer of Pharlap, and the most famous "THE Andrew Schulman Programming Series" at the time. For example, Windows INTERNALS written by Matt Pietrek is also a sensation. Pharlap is the most successful software company published the DOS Extender Software.
Talking about Matt Pietrek, those who are familiar with Window Programming should rarely do not know the master character. Matt has long written under the Under The Hood column in Microsoft System Journal, which has written some programming techniques in-depth system, and is one of the three giants of WIDOW SYSTEM Programming several years ago. Matt is also a famous Window unsature tool Softice, BoundSchecker's main R & D engineer. Matt itself is also from Borland. When Matt is working to Borland, it is developed in the Turbo Debugger team. At that time, Borland's Turbo Debugger was the strongest deactivation tool under DOS, even Microsoft could not launch the deactivation tool with Turbo Debugger counter. Matt absorbed a lot of knowledge in this group and quickly became an expert in this field. Some members of the Turbo Debugger team were taken away by Microsoft, allowing Microsoft to master the core amplifier techniques of Borland, so that it can also launch a nice amplifier tool. Matt also went to Numega to become a key figure for developing Softice, Bounds Checker. Write here or not to admire Borland, because many heavy-grade software engineers in today are cultivated by Borland.
After Watcom C / C , after the DOS market, the DOS is bound to be gradually eliminated because Window has gradually be gradually eliminated, so Watcom C / C should continue to survive, and must also launch the WINDOW platform C / C development tool. About 1993, WATCOM finally launched the first Window development tool.
However, the C / C development tools launched by Watcom C / C are really ordinary. It is similar to an ancient product that is integrated with the development environment and the other three opponents. It is not a point of characteristics. As a call as its optimization compiler. So a very interesting phenomenon at the time, that is, many software will buy Borland C / C , or one of Visual C / C , Symantec C / C , and then match a set of Watcom C / C . One of the other three sets of development tools when developing an application system, and then use Watcom C / C to produce the best program code when it is shipped. After Watcom C / C has launched the development tool of the Window platform, it still attracts a group of users, although Watcom C / C is the smallest, but it also supports a day in one side. One of the four large C / C development tools. Watcom C / C is later acquired by Sybase, and becomes the predecessor of Sybase's Optima .
For me, Watcom C / C is like a white C / C army wearing plain, but has the best training.
Key moments - mfc or not
After Symantec C / C and Watcom C / C have gradually stabilized, the four major compiler decisions are gradually approaching. Before 1994, Symantec and Watcom were facing a very severe test, that is, the choice of C / C Framework.
Although Symantec and Watcom have a market with their own characteristics, one of the most important factors at the time is C / C Framework. Therefore, Symantec and Watcom must also provide a set of C / C Framework. However, this is a difficult problem for Symantec and Watcom, because the C / C Framework is occupied by Borland's OWL and Microsoft's MFC, if you want to develop new C / C Framework, then Symantec and Watcom do not Such a strong resource cannot be completed in a short time. So Symantec and Watcom must be the next decision to use MFC or OWL as their development tool C / C Framework.
In early 1993 Symantec and Watcom, Symantec and Watcom, respectively, to sign the License MFC as their development tool C / C Framework. Here, the market has formed three forms of hitting a form in the market of C / C Framework. At that time, many people predict Borland will become a lost home, because the market has become a side, and the MFC looks like a victory. At that time, the intense debate was also launched inside the Borland, discussed whether License MFC was also made as a C / C Framework, and stopped continuing to develop OWL. But later Borland or decided to continue to develop OWL, without using MFC, because Borland's C / C Technical Group believes that MFC is not OWL in the architecture or design. And because Visual C / C supports the standard support for C / C is not as good as Borland C / C , a large number of MACROs and an inequality in the MFC are used, so if Borland C / C is used to use MFC, then modifies Compiler to compile MFC. For this point I think Borland or a correct decision, because if Borland is also license MFC, then not only lost in the momentum, and when the development of MFC is completely in the hand of Microsoft, then equal The neck is in the hands of others, and it will not be able to move. Unfortunately, Symantec and Watcom did not see this, thinking that there is a Framework like Microsoft, you can do something elsewhere and Microsoft and Borland, Symantec and Watcom don't want this, decide to make the later battle After all, the PC's C / C Development Tool Market is fully launched.
Time originating in 1994, the four big king's decisive days of C / C development tools finally became more gorgeous.
OLE's split
I don't know if it is not just fortunate or Microsoft. In 1994, the eve of Borland C / C and Visual C / C , Microsoft launched the OLECT LINKING AND Embedding technology. OLE is born in order to fight against Apple's file technology and IBM OS2 Workplace and file technology. OLE allows the Window platform file to be able to embed in different applications and enable files to edit (in-place editing) in the application. To be honest, Microsoft's OLE and Apple and IBM technology are more likely, OLE is also proven to be a failure technology, but whether it is Microsoft's OLE or Apple / IBM file technology is also Failure technology did not cause huge success. Although these documents have not been successful, OLE is enough to become an important factor in the failure of Borland, Symantec and Watcom.
I still remember that OLE seems to be a temporadic fashionable function, because word files can be embedded in Excel, and you can choose this Word file, the application is immediately called to edit it, it is really I feel very magical. However, only Microsoft's apps in all software vendors have such functions, such as other manufacturers, such as Lotus, WordPerfect, etc., can not make this feature. This creates unfair competition, because OLE technology is launched by the work system manufacturer Microsoft, but letting its application departments have this technology, while other software vendors cannot get the first hand OLE technology. What is this why this is the reason why other software vendors is so fire at that time. Although the other software companies have also launched an OLE function after obtaining OLE's technical information, it is a lot of microsoft for a long time after all, and the market has lost a lot. However, I am also very strange that at the time of built-in OLE function, almost all software vendors launched in the application of several applications and after using OLE, it is very easy to get rid of the use of Microsoft applications. The program does not happen this situation, so many people think that Microsoft has concealing some technologies without letting other vendors.
Since OLE is so complex, Borland cannot immediately make this function in OWL, so there is a negative impact on the market. As for Symantec and Watcom, although the License MFC, the LICENSE is the version of the MFC 1.x, and Microsoft does not do OLE in MFC 1.x, but in MFC 2.0. The most important feature in MFC 2.0 is that Microsoft has joined 20000 multi-line support OLE code, but MFC 2.0 is limited to Visual C / C , which is the key to let other three manufacturers have suffered.
For the impact of OLE, Borland is well known to support OLes in OLL 2.5 in Borland C / C 4.5. The solution launched by Borland is OCF (Object Component Framework).
Borland has several major goals when designing OCF. These goals include: 1. How can I make OLE trivial, complex interfaces can be simplified; second, how can OLE consistency in the window environment? Third, how can Microsoft occupy a day, the land (not necessarily, and) makes the product of the OLE function. Fourth, how to make most C programmakers can enjoy OLE's functionality without being limited to OWL's programmatists. Due to the above design objectives, it creates elegant and elastic OCF. Since the OCF itself is a complete and independent Framework, it can be applied to various applications development Framework.
I don't know whether friends who have used Borland C / C also remember one of the architectural graphs of the picture below, and the following OCF model codes, these, I found the article written in 1994 to find it, it is really I feel emotion, I also recall the scene of the time, and let you recall OWL and OCF. For friends who are not familiar with OWL and OCF, it is also possible to observe the techniques and design concepts from the following figures and program codes. Basically I now look at these graphics architectures, it will find that they have no backward now, and it can be seen that the designer's skill (Carl Quinn Again).
//
// INSERT An Ole Object INTO The View
//
Void toleWindow :: CMEDitInsertObject ()
{
001 precondition (ocview);
002 Tocinitinfo Initinfo (Ocview); 003 IF (Ocapp-> Browse (InitInfo) {
004 TRECT RECT;
005 getInsertPosition (RECT);
006 setSelection (New Tocpart (* getOcdoc (), InitInfo, RECT);
007 ocview-> rename ();
008 InvalidatePart (Invview);
}
}
Project 1 OWL TOLEWINDOW Support OLE insertion of member functions
//
// Handle Left Double-Click Message
//
Void tolewindow :: evwButtondBlclk (uint modkeys, tpoint & points)
{
PRECONDITION (GetOcdoc () & getocView ());
ToleClientDC DC (* this);
DC.DPTOLP (& Point);
TOCPART * P = getOcdoc () -> getParts (). Locate (Point);
IF (MODKEYS & MK_CONTROL) {
IF (p)
P-> Open (true); // ctrl key forces open editing
}
Else {
SetSelection (P);
IF (p && p == getocView () -> getActivePart ()) {// resync the Active Flag
P-> Activate (FALSE);
}
GetoCView () -> ActivatePart (P); // in-place activation
}
}
Cheng Style 2 OWL ToleWindow Support Left button Double-click Member Function
Although Borland joined the OLE support in OLL 2.5, helpless Microsoft added many other features in OLE, so that OCF could not fully support OLE, Borland could not retain Borland C. / C launched, so in 1994, Borland finally launched a 4.5 version of the battle.
The final cliff to the C / C development tool
"Although I have passed for a long time, I still can't forget the most tragic battle!"
In 1994, 1995, Borland finally cleared all the problems in Borland C / C 4.0, and also developed the most stable, fastest Borland C / C 4.5 version of Borland C / C 3.1, ready And Microsoft finally fight. I still remember that many books on Borland C / C and Microsoft C / C were used in the book market, and the series of Borland C / C was blue-colored, and Microsoft is With red as color, it seems that the two major legions will eventually fight.
C / C four big kings Decisive Boys Borland Lord will -borland C / 4.5
However, this battle is not only the Borland's Blue Order and Microsoft Red Army relative to the army. After Symantec's gorgeous army, after the military, Watcom's white hard-driven pillow, but also from Microsoft License after MFC, blue, red, flower, The days of the battle of the White Four Cities finally arrived. First, after Symantec and Watcom have obtained the MFC, Symantec has launched the version of C / C 7.x, and Watcom C / C menstruates. Two C / C Framework products in the same door are not easy, and the new version of Borland C / C 4.5 and Visual C / C has also joined the most important battle. But let Symantec and Watcom C / C are surprisingly surprisingly the MFC used by Microsoft is actually higher than their version (1.x to 2.x), and the new version of MFC contains complete OLE support capabilities. Although Borland also has OCF, it still does not invincibrate the OLE capacity in the new MFC. Since almost all applications are required to support OLE, only the latest version of Visual C / C can develop full OLE capabilities, so no matter whether it is in the end, let's join again. Therefore, the situation in the market has happened huge changes, almost most of the application development because OLE is selected to use Visual C / C , Symantec and Watcom Legion quickly defeat. As for Borland C / C 4.5, although it is a first-class product, if there is no OLE factor, Visual C / C new version is really better than 4.5. Although 4.5 also has OCF, only Borland and Novell, WordPerfect, choose to use OCF, after nearly one year of Visual C / C , and MICROFT MFC 2. The X version is really a stronger. Basically OCF's architecture is really a good thing, just OCF can't fully support OLE, because OLE development is in Microsoft hands, so although the structure of OCF is good, it is difficult to function. Microsoft combined operation systems, development tools and applications are really unfavorable. This is true for defeating Lotus, Borland, and is the same.
For Symantec and Watcom, this battle is like the "Battle of Changping", Qin Jun is killing more than 400,000 Zhao Jun. Killing Symantec and Watcom's full army, defeated, to this Symantec abandoned the PC's C / C development tool market, and began to develop Java development tools, and then launched a famous Visual Cafe later, as for Eugene Wang left Symantec, since then, the field of PC development tools is also left.
Watcom is even more miserable, the entire company gradually micro, and the development tool of the Window platform is defeated, and the two are empty. Soon after WATCOM was emerging Sybase mergers and acquisitions disappeared in the fierce competition.
The cause of Symantec and Watcom failed was that the Framework MFC of C / C was in Microsoft hand, and Microsoft actually held the MFC tool updated by Symantec and Watcom, and improved integration in Visual C / C . After the development environment, the key features of Symantec and Watcom appeal were completely blocked by Microsoft. Therefore, in the case of product, technology, market and momentum, it is naturally only to be slaughtered. For Borland, although there is no harmony like Symantec and Watcom, it is also a deblred. Although it is confirmed by Borland C / C 4.5, it is indeed a very good product. Whether there is a first-class performance in OWL, optimization compiler, integration development environment, and Borland C / C 4.0 is compared, there is a normal blem, and now Borland C / C 4.5 is still one of my favorite versions. However, helplessness, Borland C / C 4.0 gave people a negative impression, and unable to fully support OLE technology in full swing, therefore lost in the decisive battle. Fortunately, the blue Borland army is a well-trained, although it makes Microsoft accounts for more than 50% of the market, becoming the boss of C / C development tools, but Borland still has more than 30% of the market, a little breathing And support Borland to maintain the company's operation after the failure of important campays, waiting for Delphi's fire to rebirth, and then start again.
After this, Microsoft finally cleared most of the opponents. For Microsoft, the war of speech development tools has ended, which is destined to be mostly marketable by Microsoft. After Microsoft Hand Work System, Office Software, and Development Tools three major profit margins, Microsoft has also begun to assign the spearhead to two competitive objectives, connected libraries, and dominance development tools. After Microsoft officially entered these two markets, of course, there was also a good drama, especially in terms of the dominant development tool, opened VB, PowerBuilder, Gupta / Centura, and Delphi in terms of architecture development tools. Another unexpectedly open war is Microsoft's browser war in 1995 and Netscape.
For Borland, after C / C finally, basically I think the junction of the development tool has ended, Borland also officially started to walk down. More seriously, I think that Borland has lost the C / C Jiangshan, but also lost the creativity of C / C development tools. This is where I feel the most regrettable. I still think that Borland has not raised it. At the beginning of the Borland C / C 3.0 / 3.1, the spirit of C / C creative winds. Maybe, maybe, you have to look at C / C for Kylix or C Builder 6 can rely on this lost spirit, don't let me down.
Bika number a year of C / C World Champion-Orland C / C 3.1- forever miss
C / C development tools that never climate-IBM Visualage C / C
The role of IBM played in C / C development tools has always been a ridiculous moment, because in the C / C compiler war's most fierce moment, IBM's global information big factory has been absent. After that, after 1995, the C / C compiler market has been slowly added to the War, launched VisualAge C 3.0, attempting to attack this market. However, the market has long been said by Microsoft Visual C / C . Although the IBM Visualage can define the relationship between the object with innovative visual designers, but in other aspects, the whole integrated development environment is slow, such as snails, requires very high-end hardware to be implemented smoothly, and Visual Tools such as C / C and Borland C / C are like dinosaurs, so there is almost no response in the market. After IBM launched VisualAgE 3.0 did not receive any significant results in the PC's C / C development tool market, IBM has set a lot of resources again, developed the next generation 3.5 version, hoping to have a certain ratio in this market. I know that IBM bets a lot of resources in Visualag, because IBM from the Beta version starts, some people contact me, I hope I also write some articles for VisuaLage 3.5 on run! PC. So in June 1996 I wrote a compament article of the C / C compiler, the following information is compared to the VisualAge 3.5 and other compilers at the Beta version of the year ago:
From the above data, it can be seen that the performance of VisuaLage 3.5 is not bad, but I am not running not too VisuaLage 3.5 for me when I use AMD DX4-100 / 32M RAM machine. Later, Taiwan IBM was responsible for VisuaLage product managers asked me to eat. In this dinner, this company also invited He Yuan (later for the president of the credit), Xue Xiaoyu (the vice president of the credit) and the other two The author, I hope everyone will write to VisualAge 3.5 in the computer magazine, together with Promote this product. In this dinner, I met with He Yuan, Xue Xiaoyu, and He Yuan had a technical column in Chinese PC Magazine. I was almost unable to run Visualage 3.5 to this time, and he also immediately agreed to borrow a PC at the time IBM at the time, while writing a VisuaLage article, except for Run! PC original Outside of the draft fee, IBM will pay a draft fee of 2.5 yuan. Hey, IBM is really a big hand, I am calculated that my capacity at the time, write an article to earn 2 to 30,000, there is a free maximum machine, it is so good. But then I still think that IBM may not have deep tillage in this market, and don't agree with the concerns that don't want to violate their writing habits and defense Borland. Now I think it is quite stupid that it is good, and the contribution is not earned.
The reason why IBM's C / C development tools cannot be successful in the market is because it doesn't understand what users do in the fierce market in this competitive market. Another reason, because IBM does not have an important career in the development tool software on the PC, even if it is unable to compete with IBM, it is nothing to do, unlike Borland this can be the battle of life. So IBM just rises to play, then let go. So I think the tool used IBM in the PC platform is very dangerous because IBM may give up this market at any time. For example, I don't know how it is now VisualAge C / C , is still 3.5 or version 4.0, has not had any maintenance and improvement in several years.
Later IBM In order to launch a Basic tool that can be packed with Microsoft on the OS2 and WINDOW platforms, secret Basic has developed an Object Basic. I have also seen this tool, but Object Basic runs slowly as the turtle. Later, I didn't know if I have never been able to improve this problem. Therefore IBM has never launched this product. Now IBM seems to be interested in Java. Visualage for Java is also a good development, I hope that there will be IBM's attitude toward Visualage for Java. Like Visuaage for C / C and Object Basic. Quick Falling Potential Star - Sybase C / C Rad Tool Optima
In 1996, Sybase acquired Watcom, finally launched a C / C development tool, Optima . Optima is the first RAD C / C development tool that combines Watcom's optimization compiler and the Towed development environment similar to Delphi, and better is the Optima component architecture (similar to Delphi's VCL) is completely pure C / C program code writes. This is not allowed because this represents Optima is a super Saiyan tool that combines Visual C / C and Delphi two major kings.
After I know this tool, and after I have actually used, I am very shocked. Because this tool is to use C / C 5, 6 years, it is also attractive than Delphi. Therefore, I immediately introduced this unbelievable tool at Run! PC. Sure enough, Optima quickly started the wind roll market, although there was no immediate occupation of a large market, but has caused a momentum that has been stressed for Visual C / C and Delphi.
I remember that the product issued by Taiwan Sybase Office also attracted hundreds of people, and it was not always. After publishing in my run! PC article, Taiwan's Sybase immediately and my company. The residual governor and I met at the time, and I hope that I will continue to write articles for Optima , and Taiwan Sybase also provides an additional word plus 2 yuan. But I told the balance of rationality, although Optima 1.0 is great, but still have some bugs, and conflicts with the Chinese environment, unable to handle Chinese, and need to solve the problem immediately to success in Taiwan. She promised me immediately to react to the head office. I also honestly telling her that I can't write some uncertain things before the problem is not solved. Later, Taiwan Borland's general manager was also looking for me to ask about Optima , I told him that Optima is a good thing, but Chinese has a problem. If the Chinese problem can be solved, then it will have a big impact on Borland's products. At that time, I didn't know that Borland had begun to develop C Builder due to Optima .
In the end of 1996, Optima 1.5 finally entered the stage of Beta, but it was still very disappointed when I got Beta, because Chinese problems still did not resolve. Later, Taiwan Sybase was looking for me again. This time, I met Mr. Guo Jun, General Manager of Taiwan, and Sybase's Singapore's president, but I forgot the name of this gentleman. After we met, I immediately told them about the issue of Optima 1.5 Chinese and many bugs, I hope they can solve it, so Optima 1.5 can be successful in the Chinese market. But I'm unexpected, they don't seem to have rushing these questions, but inquiry if I have a willingness to work for Sybase, do the product manager of Powerbuilder. Maybe because I wrote too much thing for Delphi, let Powerbuilder have been greatly influenced, so they want me to work to Sybase to combat Delphi and Promote Powerbuilder. At that time, the treatment conditions were very very, very attractive, twice as much as I was higher than the salary (I was working at the policy meeting). But because I have no interest in PowerBuilder, I tell them if it is Optima product manager, then I will consider and accept it.
I didn't think that Sybase's Singapore's president told me that Optima may stop after 1.5, because Sybase wants to move the resource to develop a new Java RAD development tool to the increasing java, that is, later Powerj. So he asked me if I didn't want to do Powerbuilder's product managers, was it willing to do Powerj's product managers? Since I already knew that Borland started the development of Open JBuilder, and my interest in Open JBuilder is much greater than Powerj, so there is no Sybase. Sure enough, after Optima 1.5 is launched, not only Chinese problem is not solved, Sybase has not continued to develop Optima . A product so potential disappeared this, it is unfortunately, Optima should have a good chance to succeed, I believe that if Sybase knows that the Sybase later results will not give up Optima . The C / C RAD tool has always been to the later C Builder to complete this dream, and it is the success of Borland to enter this tool market.
C / C development tools are here to tell a paragraph, although it comes to Borland to continue to launch Borland C / C 5.0 but the quality is not good enough, the market reaction is not good, and later Borland finally announced after Borland C / C 5.02. The development of this product, Borland C / C 's glorious history has been there, it is really sigh, and Visual C / C has no opponent in the C / C development tool market. However, there is no competitive market that will make people relax, and later Visual C / C progress has become more and more smaller, the MFC has not made a big progress in the year, not as at that time and Borland C / C , every version has every version. A large improvement. It seems that the oligomed market is really bad.
As promised - Li Wei
Edited by - gordon Li on 11/01/2001 18:53:06
