Design of the New Operating System COS
A few thoughts on operating systems:
After looking through a large amount of materials, I found that a simple operating system is actually quite easy to write, and there are many experts who could complete one independently (of course, I don't have that ability yet). That is, it can boot from a floppy disk and carry out some simple disk operations. Of course this has great historical significance. But it does not have much practical significance, because an operating system is meant to be used. Only an operating system that can truly be put into practical use has value for us. Therefore, the value of an operating system is something every programmer should think about before writing it (of course, the developers of linux did not think at the beginning that it would succeed). The current mainstream is the Windows operating system, and linux is also pressing forward bravely. The DOS we advocate seems to be heading toward a dead end. Because everyone uses the Windows operating system, and in order to maintain the best compatibility and convenience, a kind of inertia has formed, making every one of us, especially beginners, have no choice but to accept the baptism of Windows. Indeed, the graphical interface of Windows makes it extremely easy for beginners to get started, while experts, especially those who came from the DOS era, remain deeply attached to the DOS character interface. DOS has low-level disk operation capabilities that Windows cannot have. So whether we can develop an operating system is not really the most necessary question. The key lies in whether we can develop an operating system with real practical value.
Our goal is to design an operating system, not necessarily to really write it out, but to put forward many advanced ideas and build the model of our ideal operating system, so that we are not behind others from the very beginning. Only advanced ideas have vitality. Our model should not only absorb the strengths of many operating systems, but also create on that basis. The ease of use of Windows's GUI interface, the simplicity of the DOS command prompt interface, the openness of linux source code... taking the strengths of a hundred schools and fusing them into one. This is our policy.
A few suggestions
1: The reason the DOS era is so unforgettable lies in a kind of art of operation. Because of the lack of hardware resources, programmers had to use their programming art so that software would use not only fewer resources but also have more functions. But the birth of the Windows system changed our programming concepts, no longer treating simplicity as a virtue. So-called software became a piling-up of all kinds of code, which, although making programming easier and closer to the masses, also greatly wastes hardware resources and brings many unsafe factors. So the programs we see now are getting larger and larger, but their functions do not necessarily increase in proportion. This makes us have to keep upgrading our computers for some so-called high performance. This has pushed us into a vicious circle: the better the computer, the less sufficient it feels! Therefore, reviving the virtue of DOS simplicity will not only improve our programming technique and art, but also greatly save social resources.
2: DOS drivers are a difficult problem for us beginners, and even experts are helpless with new devices. Some DOS drivers are very easy to install; you only need to execute one command. But some require manual allocation and configuration of hardware resources. Windows solves this by using plug & play, which indeed greatly improves hardware compatibility and ease of use, but at the same time also brings many negative effects. The operating system has to load a large number of unnecessary drivers, wasting hardware resources. If we were to load a large number of drivers in order to obtain broad compatibility, that would go against the virtue of DOS simplicity, and most importantly we do not have that much energy or financial resources to put it into practice. Many people favor Windows precisely because the Windows operating system supports many hardware devices and driver installation is simple. The unshakable dominance of Windows lies to a great extent in the fact that most hardware manufacturers will develop drivers compatible with Windows. Our effort should be "taking what is already there." DOS should be able to directly read and use Windows drivers (perhaps even universally for 95, 98, me, xp). I wonder whether this technology would be very difficult?
Of course there is another line of thought: the EFI system. This is a new BIOS standard. This BIOS not only has all the functions of the original BIOS, but can also integrate drivers. If this becomes a standard, Windows will not have such a great advantage in the future. The system and hardware drivers could then be completely separated.
3: Design of the operating interface. The GUI form of Windows is very convenient, but it also has defects, and those defects are just what DOS can solve, though not perfectly. Many people have played CS (Counter-Strike) and a whole series of 3D games, right? They have a console function that can solve many thorny problems. If the operating system adopted this kind of organization, or improved on it a bit more. Although Windows has a DOS window, it always seems very cumbersome, and can only do rather shallow functions like batch renaming and network checking; it cannot fully embody the advantages of DOS. The console method is similar to DOS + Windows 3.2, only with much greater flexibility.
4: Improvement of the DOS command prompt. Adopt a so-called semi-GUI interface. While entering commands, there would at the same time be command prompts, similar to the drop-down menu in the IE browser address bar. Like the drop-down menu that appears when writing linked lists in VC++6.0. This is very convenient, and there is no need to painfully memorize so many commands and parameters.
Maybe I can only write this much for now; I'll revise it later. Everyone is welcome to write your own suggestions after this. And mark it at the end with "Ver XX.XXX so-and-so".
How about giving this model a name? Please offer more opinions. For now it is called COS (CHINA OPERATING SYSTEM), using the DOS system model as a basis for improvement from the inside out.
Ver 0.001 Longhr
A few thoughts on operating systems:
After looking through a large amount of materials, I found that a simple operating system is actually quite easy to write, and there are many experts who could complete one independently (of course, I don't have that ability yet). That is, it can boot from a floppy disk and carry out some simple disk operations. Of course this has great historical significance. But it does not have much practical significance, because an operating system is meant to be used. Only an operating system that can truly be put into practical use has value for us. Therefore, the value of an operating system is something every programmer should think about before writing it (of course, the developers of linux did not think at the beginning that it would succeed). The current mainstream is the Windows operating system, and linux is also pressing forward bravely. The DOS we advocate seems to be heading toward a dead end. Because everyone uses the Windows operating system, and in order to maintain the best compatibility and convenience, a kind of inertia has formed, making every one of us, especially beginners, have no choice but to accept the baptism of Windows. Indeed, the graphical interface of Windows makes it extremely easy for beginners to get started, while experts, especially those who came from the DOS era, remain deeply attached to the DOS character interface. DOS has low-level disk operation capabilities that Windows cannot have. So whether we can develop an operating system is not really the most necessary question. The key lies in whether we can develop an operating system with real practical value.
Our goal is to design an operating system, not necessarily to really write it out, but to put forward many advanced ideas and build the model of our ideal operating system, so that we are not behind others from the very beginning. Only advanced ideas have vitality. Our model should not only absorb the strengths of many operating systems, but also create on that basis. The ease of use of Windows's GUI interface, the simplicity of the DOS command prompt interface, the openness of linux source code... taking the strengths of a hundred schools and fusing them into one. This is our policy.
A few suggestions
1: The reason the DOS era is so unforgettable lies in a kind of art of operation. Because of the lack of hardware resources, programmers had to use their programming art so that software would use not only fewer resources but also have more functions. But the birth of the Windows system changed our programming concepts, no longer treating simplicity as a virtue. So-called software became a piling-up of all kinds of code, which, although making programming easier and closer to the masses, also greatly wastes hardware resources and brings many unsafe factors. So the programs we see now are getting larger and larger, but their functions do not necessarily increase in proportion. This makes us have to keep upgrading our computers for some so-called high performance. This has pushed us into a vicious circle: the better the computer, the less sufficient it feels! Therefore, reviving the virtue of DOS simplicity will not only improve our programming technique and art, but also greatly save social resources.
2: DOS drivers are a difficult problem for us beginners, and even experts are helpless with new devices. Some DOS drivers are very easy to install; you only need to execute one command. But some require manual allocation and configuration of hardware resources. Windows solves this by using plug & play, which indeed greatly improves hardware compatibility and ease of use, but at the same time also brings many negative effects. The operating system has to load a large number of unnecessary drivers, wasting hardware resources. If we were to load a large number of drivers in order to obtain broad compatibility, that would go against the virtue of DOS simplicity, and most importantly we do not have that much energy or financial resources to put it into practice. Many people favor Windows precisely because the Windows operating system supports many hardware devices and driver installation is simple. The unshakable dominance of Windows lies to a great extent in the fact that most hardware manufacturers will develop drivers compatible with Windows. Our effort should be "taking what is already there." DOS should be able to directly read and use Windows drivers (perhaps even universally for 95, 98, me, xp). I wonder whether this technology would be very difficult?
Of course there is another line of thought: the EFI system. This is a new BIOS standard. This BIOS not only has all the functions of the original BIOS, but can also integrate drivers. If this becomes a standard, Windows will not have such a great advantage in the future. The system and hardware drivers could then be completely separated.
3: Design of the operating interface. The GUI form of Windows is very convenient, but it also has defects, and those defects are just what DOS can solve, though not perfectly. Many people have played CS (Counter-Strike) and a whole series of 3D games, right? They have a console function that can solve many thorny problems. If the operating system adopted this kind of organization, or improved on it a bit more. Although Windows has a DOS window, it always seems very cumbersome, and can only do rather shallow functions like batch renaming and network checking; it cannot fully embody the advantages of DOS. The console method is similar to DOS + Windows 3.2, only with much greater flexibility.
4: Improvement of the DOS command prompt. Adopt a so-called semi-GUI interface. While entering commands, there would at the same time be command prompts, similar to the drop-down menu in the IE browser address bar. Like the drop-down menu that appears when writing linked lists in VC++6.0. This is very convenient, and there is no need to painfully memorize so many commands and parameters.
Maybe I can only write this much for now; I'll revise it later. Everyone is welcome to write your own suggestions after this. And mark it at the end with "Ver XX.XXX so-and-so".
How about giving this model a name? Please offer more opinions. For now it is called COS (CHINA OPERATING SYSTEM), using the DOS system model as a basis for improvement from the inside out.
Ver 0.001 Longhr