China DOS Union

No. The DOS kernel should be developed. Isn't LINUX just a rewritten UNIX?
Only by solving the problems fundamentally can it be kept from disappearing.
First there's the memory problem, then job scheduling (preemptive multitasking?), and also support for file systems and networking.
There are too many problems, and even if all of these are solved, compatibility must still be considered, and the internal and external commands and config files would have to be greatly expanded.
Then it would become very hard to use...
But I believe there must be a way to solve it

OP, don't speak so absolutely. What kind of applications there are determines what kind of system will be used.
If it's only putting up a termiator server, I don't like using unix or windows, but DOS instead.
Because for that kind of server, DOS is very good and very stable.
But if it's a DataBase Center, I would choose unix.
If it's making a simple interface, I like windows.
The vitality of dos lies in the fact that we are still using compute for many simple-purposed (not simple) applications

I'm facing exactly this kind of problem. A DOS program I want to write must support protected mode, multitasking, Socket, Web publishing, and FTP Server. Hehe, it's hard.
Can anyone give some suggestions?

“A DOS program I want to write must support protected mode, multitasking, Socket, Web publishing, and FTP Server”
Then that's not dos, that's unix

REAL/32 DOS(Multiuser DOS) is exactly a DOS that is multi-user, multitasking, and 32-bit protected mode, and it also supports FAT32 and large hard disks, etc. The only drawback is that it isn't GNU GPL, and that is exactly what FreeDOS makes up for.

Damn! That's so well said!

Linux+Windows+BeOS+PALM=NEW DOS. This is my idea. I think a good OS first needs the ease of use of WINDOWS and the multimedia performance of BEOS, then it needs the stability of LINUX, and in addition it should probably also have the possibility of being developed into a mobile phone operating system. Right now LINUX and WINDOWS, Palm OS are all fighting over the mobile OS territory, and that is the future trend of electronic technology. If DOS is to continue becoming stronger, it must have its own piece of ground in mobile OS!

Hei Xia's insight is very high, I very much agree with your opinion!
If DOS is to develop, it must rely on rewriting the kernel, while having major expansion\enhancement on the basis of preserving its original compactness and flexibility.

The reason DOS is DOS, and not UNIX, Linux, Vxworks, or Windows, is because it has its own characteristics. Single-tasking can be said to be a weakness of DOS, but from another angle it is also its strength. DOS has its own strong points, such as direct operation of I/O ports, fast interrupt response, fast startup speed, etc. Personally I don't think DOS should be compared with the strengths of other OSes using DOS's weaknesses. What fits is best! DOS's strength is not in home or personal applications, but in the professional application market. For example, in the embedded industry, there you will find that it is a vast world for DOS.

Developing GUI programs under DOS is not difficult. Some special-purpose programs can be developed. For example, if you have old computers and some important materials on hand, it would be entirely possible to develop a complete special-purpose materials system, using DOS to manage the files on the system, together with specially written DOS programs to manage the materials. Its work efficiency and stability would definitely be better than special-purpose programs under WINOWS (for example, electronic maps, library book search systems, and shopping mall cash registers are all good ideas). Or you could also burn the materials, programs, and DOS together onto a CD; no matter what machine it is, as long as it can boot, it can read the materials.
If you insist on developing those multitasking graphical interface programs, and also keep supporting new hardware and technologies, then it really is not an easy matter. At most, it would only earn a fine reputation. Besides, even after so many years of development in this field by MS, WIN9X was still full of BUGs, and in the end they gave up and turned to developing the NT system. Therefore, practicality is the most important factor in DOS's survival.

I think it is very important to develop an operating system core that supports double-byte characters

I think DOS has its own unique advantages in developing embedded products: low development cost, short development cycle, and fairly abundant resources. dr－dos and free－dos are free, and multitasking operating systems and TCP/IP protocol stacks are also free and provide source code. If everyone is interested, you can download the source code for ERTOS and TCP/IP from the following websites. They are all free.
http://www.wattcp.com/
http://www.ertos.com/index.shtml

I've been using DOS since the early 1980s, and have developed applications and devices on XT, AT, 386, 486, and Pentium. Right now I mainly work on embedded PC modules, and of course they are DOS systems again. As I see it, for the overwhelming majority of applications I am currently facing, DOS single-tasking is already sufficient, including providing powerful network communication capabilities such as TCP/IP, WebServer, and GPRS/CDMA Internet access. I think the main problem at present is development tools. We recommend that customers use BC3.1, but young engineers lack understanding of DOS, and their development speed is often much slower than expected. If there were a Windows-style IDE similar to VC, it would definitely speed up DOS application development. I believe that only when something is truly put into use can its vitality be shown. I would very much like to get opinions from DOS experts, especially on the subject of development tools. Thank you. Our company's website is www.emtronix.com, please give us your advice!

I've been using DOS since the early 1980s, and have developed applications and devices on XT, AT, 386, 486, and Pentium. Right now I mainly work on embedded PC modules, and of course they are DOS systems again. As I see it, for the overwhelming majority of applications I am currently facing, DOS single-tasking is already sufficient, including providing powerful network communication capabilities such as TCP/IP, WebServer, and GPRS/CDMA Internet access. I think the main problem at present is development tools. We recommend that customers use BC3.1, but young engineers lack understanding of DOS, and their development speed is often much slower than expected. If there were a Windows-style IDE similar to VC, it would definitely speed up DOS application development. I believe that only when something is truly put into use can its vitality be shown. I would very much like to get opinions from DOS experts, especially on the subject of development tools. Thank you. Our company's website is www.emtronix.com, please give us your advice!

Since so many people like DOS, why not develop it into something like linux? The sea accepts all rivers; thus it becomes great!