Meet the Linux Guardian

Article Title: Meet the Linux Guardian. Linux is a technology channel of the IT lab in China. Includes basic categories such as desktop applications, Linux system management, kernel research, embedded systems, and open source.
Author: Chen Hairu
　　
The vigorous development of Linux is the result of the endless struggle of countless well-known and unknown people in the world for a common ideal. Among the countless and respectable people, there is a special position. Here, he is now called "the guardian of Linux" because his job is to maintain the Linux kernel.
　　
As a fan of free software, becoming the guardian of Linux kernel is a dream. Marcelo Tosatti is a developer of Conectiva, a Brazilian Linux publisher. He is currently the maintainer of the "stable" Linux kernel version 2.4. His predecessor was Alan Cox, a famous character in the Linux community. As a Linux user, we should not forget these guardians when enjoying the endless fun of the powerful Linux. So as a guardian, what is the real side of a mysterious and unattainable person in the eyes of many people?
　　
Tosatti tells us about our daily work and other things that everyone cares about. Let's stay in touch with the Guardian of the Linux kernel.
　　
Q: Which of the following is more difficult for you to be a programmer and a kernel maintainer?
　　
A: Of course, it is easier to be a programmer. Because I used to be a programmer, I generally want to do it, or I have to do it. I only need to work in a certain field of the kernel. Now, I have received emails about VM (virtual memory), ACPI, and driver. There are too many such things. Therefore, I have almost no idea, because I have to deal with many different things. This is indeed a very difficult job.
　　
Q: Do you think your current job is different from that of nine months ago?
　　
A: I think the pressure is much lower. At the beginning, I was not used to various interviews and meetings, but now I am used to it.
　　
Q: Do you still have time to work for other projects?
　　
A: I think, but I cannot, because I cannot be distracted. I need to systematically complete a lot of monotonous work, although this will make people a little boring, but it is very effective.
　　
Q: Is it not the same as developing a Linux 2.5 kernel?
　　
A: Of course, development is more attractive.
　　
Q: Why do you want to maintain the kernel?
　　
A: I don't know. maybe I want to have some new experiences and try some new things.
　　
One day I received an email from Linus, and I was surprised. This is a message from Linus to Alan. Linus said to Alan, "you don't want to maintain the kernel. please recommend someone to me ." Alan said: "Marcelo or Andrew Morton ." "Yes," Linus said ." So they sent me a letter. I agree, that's simple.
　　
Q: Do you realize that you have to do a lot of work?
　　
A: When I send an email back to Linus? No. Now I find that I have to do a lot of work, but I think it is worthwhile.
　　
Q: Alan Cox has been working on Kernel Maintenance for many years. Do you think you can do it for a long time?
　　
A: No, I don't want to take a long time. I don't know if Linus will continue to let me maintain version 2.5. I think he will probably. However, I only want to maintain version 2.6. Then we will do some other development work.
　　
Q: What are the greatest challenges of this job?
　　
A: Make the right decision. This is indeed very difficult. I mean, some problems do not seem very obvious. Sometimes you find an error and you have fixed it, but it sometimes causes many other problems, so you must be very careful.
　　
You must be responsible for maintaining a stable kernel. Adding various codes to the kernel is very easy, but when there are countless people working for it, there are still countless people using Linux, then I realized that I had to be very careful. In my opinion, this is a challenge, at least for me.
　　
Q: What are the main problems with this kernel?
　　
A: VM may have some problems when running on high-end machines. In addition, there are also issues with the IDE, and some chipset may damage data. I am very worried about this. Therefore, IDE and VM are the biggest problems in this version.
　　
Q: Do you think the kernel is getting bigger? in other words, the kernel expansion will become a problem?
　　
A: In a sense, yes, but it is not a big problem. If you have fewer drivers, that would be fine. In fact, there are usually a lot of drivers. You have a core and a driver, so if the driver expands, it's not a problem. All we have to do is ensure that the core part is not expanded.
　　
The number of drivers will definitely increase, which is a problem. Because, for example, when you change the API, you must change all drivers. We have to deal with all the drivers, because we are dealing with a large number of hardware. Q: What kind of research do you think makes you proud?
　　
A: Research on VM performance and on VM and processor.
　　
Q: Will you continue to engage in research and development in the VM field?
　　
A: No. I have been working in this field for a long time, so I want to change to another field. If I have time, I will do some SCSI subsystem and some research on high-end and high-availability. But I don't think I will have time.
　　
Q: There are many potential personal attacks in the Linux world. How do you deal with this situation?
　　
A: I try my best to consider the technical issues. I think this is very important. In fact, if we do not pay attention to the way we speak, it is very easy to cause quarrel. For example, I usually say, "I don't think you are right ." Without saying: "You stupid ." However, I think this is not a big problem for me. I just need to pay attention to using appropriate words and be friendly to people.
　　
Q: In fact, it is generally a technical issue. how do you think there is a conflict?
　　
A: people engaged in this line are generally at your own risk.
　　
Q: What do you think is the greatest opportunity for Linux?
　　
A: We have already entered many market fields. The embedding field is a very large market, and I think many people have not really realized how big the market is. In fact, the Linux operating system installed in your camera is likely to be everywhere. So this is a very large market. In my opinion, this is also a great challenge, because it is very complicated and difficult to satisfy various needs.
　　
Enterprise applications are also a big market and a good opportunity for us. Linux is going deep into enterprise applications, but in the desktop field, the development of Linux is still relatively slow.
　　
Q: Some people who oppose Linux believe that Linux is full of loopholes and patches are continuously applied. What do you think of this idea?
　　
A: I disagree with this idea. In fact, it depends on what you want to fix. we generally issue small patches. Linux patches are frequently released. This shows that we are constantly working hard, because no software is perfect and there will always be errors, the patch release demonstrates that we are constantly improving. If Microsoft is the same as us, I think their update frequency will be higher. Because their software is much larger than ours. So I think these anti-Linux arguments are purely nonsense.
　　
Q: Red Hat, SuSE, and other publishers have begun to release some server versions to enterprises, which are rarely updated and easier to control. What do you think of this practice?
　　
A: I think this is necessary because enterprises do not want to make too many changes. they do not like to use patches to update their systems frequently, therefore, it is necessary to release some relatively stable versions.
　　
Q: Your company Conectiva recently joined forces with three other Linux publishers to change the Linux Released to UnitedLinux. MankrakeSoft, however, criticized the move and held that it was not for technical considerations but to consolidate its own market. What do you think?
　　
A: This is a commercial issue. it is indeed a purely commercial issue. However, this method is good because it reduces the number of large commercial releases.
　　
If there are too many releases, authentication is a big problem. For ISVs, to prove that their software can run on all releases, the cost will be considerable. For example, Oracle must test that their software can run on all versions. Therefore, unification can make ISVs easier.
　　
Q: What does MandrakeSoft imply that publishers and ISVs only need to prove that they comply with the LSB (Linux Standards Base) and do not need to authenticate UnitedLinux, Red Hat, or any other releases?
　　
A: This is unrealistic. Because many things are not covered by LSB, you cannot only authenticate LSB.
　　
Q: What challenges do you think Linux will face in the future?
　　
A: let ordinary people use Linux. In fact, people do not use Linux because it is not difficult to use, but mainly because of habits and conceptual inertia. For example, some people say that Linux installation is too difficult, but in fact, if you allow many people to install Windows, not all users can complete the installation.