Home Closet The perfect operating system! Is it possible?

The perfect operating system! Is it possible?

by admin

On the Internet, the most beloved and discussed topic of discussion is the hooliwars. It is clear from the name itself that these are wars because of "religious" leanings. It is said that truth is born in debate, but this is absolutely not true. Very often the disputants are "religious" fanatics and they don’t try to determine the real best side, but simply smear their opponent and the symbol of their "religion".
The most fertile ground is the operating systems debate. Can we stop them once and for all and make a perfect system? Let’s try to find out
It’s very common in the cholivars to put down Windows (I do it often and with pleasure myself). But many people are not willing to give it up just because of a small set of applications they really need for work. So, as a kernel of an ideal system, you can use anything, but it is necessary to provide support for the Windows software. There are plenty of emulators for linux (in the neighborhood of 50), have a look at comparison chart yourself! There’s emulation of anything out there, not just wind. Talked to some myself and it worked 99% of the programs and toys I personally tested. You could try to "talk" the program, but that’s not the point. Incomplete support of a program is limited only by the closedness of the Windows. It is enough to open all source code (at least by court order) and very soon software support will be full.
The ideal Operating System, hereinafter simply called OS 🙂 must support real-time operation. QNX has this implemented, you can take its kernel as a basis and screw everything else.
The OS must support all existing binary for a given processor, support ports and builds from source. In fact, this is the easiest work of all the way to the perfect system.
The OS should have a modular and configurable kernel. Linux has it, so let’s take it from there.
Make the file access system a bit deader than the Unix one and noticeably cut it from the Windows version. The group membership system should be taken from the Windows and trimmed with a coarse rasp.
You need a single directory service to store network information. AD’s implementation is ugly, so take it as a basis, throw out half of it and refine what’s left a bit.
The file system must be journalable, with full data recovery in the event of a sudden power outage. NTFS is absolutely not suitable for this, it crashes very easily. And you lose not only the underwritten file, but sometimes a whole list of directories and files. Recovery tools can recover files which will be messy. Take something like ext4, ReiserFS.
Security is the most painful issue. Take OpenBSD and twist, twist, twist. Although it is vulnerable not because of the implementation of specific programs and modules, but because of fundamental flaws in the network protocols used. Only to come up with the perfect protocol and in a short time, and even with little blood, to change it will not work. The packet filter and intrusion detection system should also be built in. And to realize this on different "rings" of the processor. In the x86 series processors, there are 4 rings; only 2 of them are used in Windows: the privileged one and the user one. In the IOS it is necessary to use all 4 rings, and the lower level is not the privileged one but the system controlling the work of the system kernel. A sort of IDS and anti-virus heuristic analyzer to protect critical system locations against malware. Even if it gets to the driver level or the kernel level, the virus should still stay a level above the protection system.
No matter how you look at it, an ideal system involves the complete opsourcing of all the OSes on the market and working together on the OS. Fans of Open Source, you’re going in the right direction! Unix support for binary of other Unixes brings them closer to the concept of an OOS. The development of more and more emulators will also bring the OS closer. The work in this direction is in progress. Gradually, "out there" or "in there", very visible features of the OS are appearing. It is enough to gradually assemble the best and the constructor will become that very IOS. Quantity will turn into quality. Picking up "features" and "tricks" from one another by various operating systems is an indication that a "bright future" is approaching. But slowly. And only Microsoft with its closed standards and other nonsense sticks in the wheels of progress and generally drags everyone to the bottom. OPSORS IS THE FUTURE!
If someone develops the topic, I will be only glad. Perhaps the topic will get the necessary development and attract the attention of specialists involved in direct work on modern operating systems. And maybe some of these items we will be able to see in the nearest distributions.
P.S. I absolutely did not intend this essay as an advertisement for Open Source. I just read a lot of chatter and pondered "aloud" about the possibility of creating an IOS. Gradually came to the conclusion that opsorsors is a ray of light, pointing in the direction of "bright future".

You may also like