Devil and Dreams of the private language ------ About the software development of the demon dialogue (1) sequence devil blows the flute: Maybe this is not a preface, just for me to discuss with Mengyan so long after a long time Feel. I wandered in the Internet for 5 years, I only encountered such an in-depth discussion of technical issues, I should only have a man. Although I didn't personally and he met him, I can see him from our exchanges. Many problems also look into a huge contrast. Everyone can experience this when you look at the following text. As for the debate, I don't think there is. We don't want to do what we do, we must not say what we said. "People have never been understood, but they often make mistakes." Karl.Popper has always affected me now, but no matter what debate, what is final, it is always necessary to rely on practical to pay for pseudo. In the following article, everyone can see our idea and constantly sharp. Change. A large part of the reason is that we have a constant certificate of our own point of view. From the first letter, I will start to study STL and the source code of the later Boost. Many of these guesses now come back to see that they may be comic, but they are still very meaningful. We discussed a discussion of GP and OO, which also involved issues from Java, COM, .NET, GC, philosophy, economics. We have to discuss far from being over, but our current discussion has slowed down the direction. "Now is not there.". "(Churchill 1942). Because it is not an exclusive communication, although we do two best to make changes but because we work is busy, there are a lot of messy places, please understand. However, I think there is still a lot of good ideas and information, but I need to rely on everyone to dig and experience it. Dream: I agree to disclose our letters. When I was just an initiator, I often saw the devil's article on the Internet. In addition to admiration, I also thought that one day, I would have a fierce argument with him. At that time, I was still a non-crux of students, and students like controversy, and I am a student who specializes in controversy. In this year, many things happened to me, and I have changed much, but I still love to talk, but slowly become less love. Occasionally, I will have a complaint between the Internet, most of which is unsolicited. Everyone has their own point of view, and the dispute between language martial arts is also good, and the dispute between design and disadvantages is also good. I am not a very smart person, not good at predicting, why should I affect someone else? Why do you have no debate? Is there any thoughts, it is going to do it, it is good, and it is also harvest. What is the meaning? At this time, this kind of mood, there is still someone who can pull me for such a long period of time and a big level, may have only de devogous brothers can do it. Devil's knowledge is broad, be good at grasping the overall situation, and there is very strong actual hands-on ability and learning ability, I admire him very much. The following letters are some of our early communications, and the time is far away, but it is now not close.
Not only because my mood and the idea of software have changed, and because of our current discussions have gradually transferred to specific design issues, I now read these previous letters. How much feels don't be prans . Published, just want to make a commemoration, and maybe it may cause some netizens' thinking, then our purpose will reach. I still don't advocate the endless debate, let alone this time, I feel the lack of practice experience, I hope to do something in the future, for programming, software's view is dramatic, please don't be because of us Some speech is worried. You may not be wrong, we don't necessarily, everyone holds a belief, do it, this is the correct attitude. *********************************************************** *** Hi Myan: I am very happy to communicate with you. It seems that you are very studying for C . I basically agreed with you: "Python's GP and STL's GP implementation is different." But I don't know why you think C is still the GP winner battlefield (of course, we rule out the C / C than Python has a wide range of factors). In my understanding, Template and GP as a programming idea and method is actually a remedial measures inherited by C . The emergence of GP has enabled OO to make it clearely and simple to portray the world. In the real world, the description can be explained in such a way that the wheel of a car has a wheel's own movement method and function, the engine has the motion mode and function of the engine, the control system has the control system's own Methods and functions, the skeleton of the car has its own way and features they are unassays. When we assemble a car, we used the car to combine these different functions in a multi-inheritance manner. OO should also be doing this as a programming idea, but C has caused more inheritance in actual use process due to multi-inheritance (such as typical diamond type structures). I don't know if you have used OWL? OWL is a multi-inherited framework class library. However, OWL has used many inheritance, which has led to instability of the entire frame. At the beginning, I will have more dizziness when I study OWL Document / View. Since then, the MFC, VCL does not have an indexed model using a single inheritance. This avoids the ambiguous disambiguation problem of inheritance to some extent. However, the benefits of inheritance have also been abandoned. If you use a single inherited method to assemble the car's ideas into engine, wheels, skeletons, and control systems are inherited from iron atoms, and the iron atom uses different forms. Transition into various components and then it can be assembled together because it is the iron atom. Such a car is made of exceptions but our design ideas become extraordinarily weird. Why do the cars do not consider from iron atoms. Bjarne invented the approach of Template and GP independent objects. The GP itself does not require support for the virtuality of the object. It is manifested in independent and single but he can abstract the essence of things. Therefore, we have a strong tool to use more inheritance methods to come from new planning our design ideas. That is, we use GP to describe the independent things motion model, then use multiple inheritance patterns to put them into the object. Tempalte and GP provide a theory of we look at wheels, skeletons, and engines, or a reason to say. So I think the commonality of GP and OO is abstraction, and differences are the independence or individuality of OO emphasizes things or GP emphasizes things. We feel that OO is tending to complicate and confusing because we have no GP this powerful tool. When we have GP, OO's defects are sympathy to solve the OO portrait world, and more clear and simple rather than far-renewal, all things will see all things as atoms. The difference between the two methods can be known to observe the difference between MFC and WTL and ATL.
Last time we discussed GP and OO to talk about philosophy, see your post thinks your philosophical literacy is still a little Popper's dialogue. I don't know how my feelings of your programming philosophy are neither science is not philosophical. If you don't have so much, it is very late. I hope to see your letter as soon as possible. Ian.king Demon Blows the flute: This is my Jan's reply, this letter is temporarily uniform. However, behind the debate is big, we discussed a little Karl.Popper's philosophy in the end of the letter. / Snowfalcon brother, hello! I am also very happy to discuss with you privately. Because the pressure of the project is now heavy, I am as simple as possible. First of all, I used to be a person with a pronounced language. Do not put any language in your eyes except C / C . But as I know more and more about C , use more and more proficient, now you can look at the language from a higher angle, no longer a language. I realized that although C is a great language, it is very highlighted, it is difficult to cultivate a sufficient number of skilled C programmers. Therefore, in the development of the programming language, maybe C roles is more like a pioneer, and may not be the final mainstream. Today, Java has begun to occupy the mainstream in the OO field. In the future, this scene will definitely repeat in the GP field. Therefore, in the future, I can continue to become the main application in the future (Application) development language is tendency to pessimism. In 1994, the Design and Evolution Of C was published in 1994, Bjarne Stroustrup pointed out that C is not intended to be a large application system development language. If you have to do this, one is to have a supporting application class library, and the other is to join GC (garbage collection) support. So why is I in the study of C so enthusiastic? Because I think C languages have the status of hub and milestones in the development of modern programming languages. Just like Kant's status in philosophy, you can praise, you can also blame, but in the future, there is no language when designing, evaluating and discussing, can bypass C . In three aspects of the process, object-oriented, generic, C expanded to the limit of a static strong type language, and also derived Template meta-programming, and has the ability to simulate Functional Programming, and its broad , Amazing, Bjarne Stoustrup as a language designer, his courage and guts, it is really difficult to surpass. Therefore, I firmly believe that if I can really understand C , then in the future, no matter which language to use, I can look at the deeper eyes, will not be confused by the appearance in front of you, will not lose the direction. In the future, I think Java / C #, which is developing, is developing D language, and of course Python, has a good application prospect. In addition, I also optimistic about the ADA 95 will have a certain development. This language is the most respectable language in addition to C .
