• 0

How to build an operaing system ?


Question

Recommended Posts

  • 0

"ReactOS is an Open Source effort to develop a quality operating system that is compatible with Microsoft Windows® applications and drivers."

"The kernel has come a long way and is quite stable and many APIs are complete enough for higher level work to be done. Substantial work has been done on many libraries. Initial networking code is there, but unusable for the moment. A basic implementation of the Graphics Device Interface (GDI) and a VGA graphics driver is paving the way for a GUI. The amount of developers is steadly increasing."

its realy nice os, and you can use losts of windows software on it

they even have a live cd version

  • 0

1- Mostly x86 Assembly and pure C, C++ was not really a good choice at that time

2-Run it under an emulator, Vitual PC and Vmware works fine, use bochs if you need a free one

3- You absolutely need know to x86 Assembly to make the Master Boot Record and some parts of the kernel. Assembly is the lowest level language you can program with if you dont want to learn all the x86 OPcodes. They are numbers that means something to the processor like making an addition etc...

4- Windows 95/98/me all relied on a dos subsystem, in Me it was simply hidden, you can even get a patch to return to DOS from Me. They just had no choice or else no one would be using windows at that time, because customers needed perfect compatibility wth DOS programs to use old softwares that was popular at that time. Window NT/2000/XP was made with stability in mind, not compatiblity, they are able to run DOS programs but in a CONSOLE which is not DOS but only program which tries to emulates. That's why a lot people complained about certain DOS programs not functionning correctly or at all with XP, XP was the first microsoft attempt to bring an NT OS to "normal" users. WinXP is just Windows 2000 with a better user interface and better compatibility with games, just look at the build numbers of win2000 & XP, 2000 is windows NT5 and XP is 5.1 . Longhorn, the next windows, will be windows version 6.0 which should be a significant upgrade from XP.

So... You want to make an OS, learn some assembly, pure C or C++ ( which is not as bad for that than it was before it was standardized ) and have lots of patience. Start by reading how to make MBR and some x86 specific stuff like gate A20, gdt, ldt , interrupts and last but not least, protected mode. If you don't know C/C++ start learning it before thinking of doing anything. And don't even think of doing a user interface before a lot of things is implemented. You actually have to reinvent the wheel, you have to access hardware like the video card directly to write something on the screen.

If anything i said is not 100% correct, just tell me in a correct way.

Thank you

  • 0
I can tell you that Windows is written in C and C++, with small amounts of assembly in places. Linux, mostly C++ from what I understand. DOS, who knows (or cares) ;)

Linux is written in C, Linus torvolds the creator has said a million times that C++ is slow because it has extra overhead. Just so you know ;)

  • 0
Linux is written in C, Linus torvolds the creator has said a million times that C++ is slow because it has extra overhead. Just so you know ;)

585816262[/snapback]

Got it (Y)

But you didn't need to revive the thread just to tell me that. You could have sent a PM or something ;)

  • 0
Wow, for such a smart guy(Linus), that sure is an ignorant thing to say.

585816478[/snapback]

But Linus is right. For an OS, you want to spend as little time in kernel space as possible and that means you don't want to deal with the overhead of C++ (exceptions, RTTI). Also, C++ uses a lot of stack space. You only have a limited stack (8k or 4k) in the kernel. Another feature of C++ that gets in the way is name mangleing(sp?). When you have to mix languages, dealing with name mangleing is one more thing you have to worry about.

  • 0
Thanks for bumping the thread. The first three pages are ****ing comedy!

585817579[/snapback]

Yes, it's extremely funny. Until you realize that these people aren't trying to be funny.

Now it starts to lose it's charm.

It starts to become disturbing. You realize that these people actually exist and that they actually think, speak and react in this manner. It's frightening to think that these people believe that they are being clever.

This does not bode well for the future of mankind. I don't think there's any legal way to stop them from spreading there malformed genes into future generations. You can't just go around shooting the idiots and dumping them into a big hole. Although I have considered it in the past. I don't think we'll get that far with that method.

We must educate the poor losers. We must somehow convince them that they should not reproduce.

Well this is far too deep of an issue for the pages at Neowin.

  • 0
But Linus is right.  For an OS, you want to spend as little time in kernel space as possible and that means you don't want to deal with the overhead of C++ (exceptions, RTTI).  Also, C++ uses a lot of stack space.  You only have a limited stack (8k or 4k) in the kernel.  Another feature of C++ that gets in the way is name mangleing(sp?).  When you have to mix languages, dealing with name mangleing is one more thing you have to worry about.

585817508[/snapback]

That's assuming you use RTTI, and program in an OO way. C++ doesn't force you to do either. Concerning mangling, extern "C" fixes that. Linus implies that you have to use all the features of C++, if he truly said what was posted. And if that is the case, that is why I think it was ignorant. I'd be surprised if Linus didn't have more to say about programming style instead of the language. You can program in an OO way in C, too.

  • 0

if your gonna make an OS.. make a good one for starters LOL

or even an OS, at your expierence level, i'd be surprized.

but lemme give you some advice...

an OS takes years and years to profect.. and you need to know ins and outs of computers

an OS is millions of lines of code. and... if you can get passed 10,000, i'll personally shake your hand.

im not trying to put you down, but look at how many employees work for microsoft. i bet more then 60% of them work in creating / improving, or in some sorta way, have their hand in the creation of an OS.

  • 0

Thanx a lot for people who posted links to learn from it, and no thanks for people who is trying to depress me. Just tell me what you know or stop posting.

Do i said that i want to make OS like windows :blink: , i just want to make an OS for my self, happy now !

*Thread redownloaded for summer holiday*

and please can some one help me in this thread :

https://www.neowin.net/forum/index.php?showtopic=311820

I think I will find my solution from Neowin.net after 10 useless posts :rolleyes:

  • 0
Not necessarily - Unix was created using C. So was Linux for that matter.

585819807[/snapback]

Correct me if i'm wrong, but isn't Assembly necessary at times, to interface with hardware? Like wouldn't Assembly be required to write some of the fundamental aspects which then a higher-level language can use off, i.e. C, C++.

Btw why the hell isn't this thread locked already?

  • 0
Correct me if i'm wrong, but isn't Assembly necessary at times, to interface with hardware? Like wouldn't Assembly be required to write some of the fundamental aspects which then a higher-level language can use off, i.e. C, C++.

Btw why the hell isn't this thread locked already?

585820040[/snapback]

The thread remains open because it is a valuable learning resource. It also demonstrates the complexity and sheer size of an operating system.

And assembly is necessary, at least for the kernel and some hardware interface. Unix was not created entirely with C code.

  • 0

first pages of this threads are jokes. but i guess he is learning. but i agree with something posted in here before: you should just do the shell for windows which will do all you want. now that would be useful and realistic. and i think it is the more important part of you final idea. and if you get your ai interface think to work nicely - that you can think of writing underluying ssytem. not before.

  • 0
The thread remains open because it is a valuable learning resource. It also demonstrates the complexity and sheer size of an operating system.

And assembly is necessary, at least for the kernel and some hardware interface. Unix was not created entirely with C code.

585820982[/snapback]

In that case perhaps you should strip away the non-informative bits of the thread if possible.

  • 0
Btw why the hell isn't this thread locked already?

.....

John:

The thread remains open because it is a valuable learning resource. It also demonstrates the complexity and sheer size of an operating system.

.....

In that case perhaps you should strip away the non-informative bits of the thread if possible.

585821443[/snapback]

:yes: :yes: :yes:

I tend to agree with Winston, this thread should be closed. maybe create a new sticky thread on building an OS?

I think we pick on Elagizy way too much in this thread to allow people to bring it to life again.

  • 0
WinXP has a command prompt, not DOS.

Just because it looks like DOS doesn't mean it is DOS.

585829659[/snapback]

correct.

earlier windows used DOS.. and i believe since windows NT, they changed to Command Prompt.

  • 0

The only DOS in Windows XP is the DOS virtual machine used for running legacy apps. This is not unlike what happens when you load Virtual PC. The DOS virtual machine thinks that it is DOS and thinks that it had direct access to the system, but really it's running in a "sandbox." When it makes a call to the hardware, Windows XP's compatability layer says "Oh, the DOS VM wants to access the hardware directly. That's not allowed any more, so I'll just give it the answer it wants and only pretend that it actually accessed the hardware."

Any non-blank autoexec files you find on your system are either:

A) Leftovers from a previous DOS/Windows 9x installation, or

B) Used to set environment variables (like the PATH statements posted before) so that programs running in the VM can find things they need.

There really isn't much more to it than that.

The Windows NT kernel bears no relation to DOS in any conceivable way. It most closely follows the design of OpenVMS (whose creators developed NT for Microsoft). In many ways it is comparable to the Mach kernel that Apple's OS X runs on top of. NT relies more heavily on threading, which makes it more scalable in multi-processor situations. However, both NT and Mach are built to be simple, versatile kernels on which you can run pretty much on any platform and support pretty much any user-space system (just like Apple runs FreeBSD on top of Mach, and Microsoft runs various "subsystems" on top of NT - including Win32, POSIX, OS/2, etc).

This topic is now closed to further replies.
  • Recently Browsing   0 members

    • No registered users viewing this page.
  • Posts

    • I think there will be a 27H1 for actual users of 26H1 The 25h2 supports ARM too : Snapdragon X, Snapdragon X Plus and Snapdragon X Elite
    • Poll: Grand Theft Auto VI price predictions, cast your vote by Pulasthi Ariyasinghe After years of waiting, Rockstar will be solidifying the launch date of Grand Theft Auto VI with the launch of pre-orders next week. While the studio has confirmed a date for this occasion, it is yet to attach a price to the highly anticipated game. So let's see what our readers think it will cost at launch. The Grand Theft Auto VI pre-orders will kick off on June 25 for digital and physical editions. Unless some last-minute changes happen, the release date will be November 19, 2026, across Xbox Series X|S and PlayStation 5. Unfortunately, there's still no information about a PC version from the developer or the publisher Take-Two. Now the question becomes, how much will Grand Theft Auto VI cost at launch? The game is predicted by some analysts to be the biggest launch of an entertainment product ever. With the amount of hype that has been built behind it and with ballooning development costs, Take-Two may price this Grand Theft Auto entry differently from other AAA titles. The current price of a AAA game is $69.99. That norm almost rose to $79.99 before calming down. But with such a massive release, Grand Theft Auto VI may be the game that pushes the boundary again. It's also possible that Take-Two keeps the price relatively low to increase the number of players that jump in early and keep them hooked on Grand Theft Auto Online to spend on microtransactions for years to come. Keep in mind that the below poll is asking for a prediction of the standard edition price, not a deluxe or any other special edition that Take-Two will introduce for additional benefits. Also, there is also the chance of the company splitting up the campaign and online portions. If you think that will happen, put your vote on what you think will be the total cost of the two. Poll Poll: How much will Grand Theft Auto VI cost? $59.99 $69.99 $79.99 $89.99 $100 or more Submit Vote If you have a very specific prediction in mind, sound off in the comments below.
    • Would you please fix your graphics. They are outdated and don't fit the article.
    • The Light of Life? We actually do glow till our Death, study finds by Sayan Sen Image by Rafael Rendon via Pexels A study by researchers at the University of Calgary has found that living organisms produce an extremely faint light known as ultraweak photon emission, and that this glow appears to drop significantly after death. The research was published in the Journal of Physical Chemistry in April 2025 and quickly drew widespread attention, leading to more than 200 news stories about the findings. Ultraweak photon emission (or UPE), sometimes called biophoton emission, refers to tiny amounts of light released by living cells as a result of normal biological activity. A photon is the basic particle of light, and researchers say every living system examined so far, including plants and animals, has been found to emit these photons. The glow is far too faint to be seen by the human eye. “I suppose it has a little to do with people being reminded of auras,” says Dr. Christoph Simon, PhD, one of the authors of the study and a professor in the Department of Physics and Astronomy in the Faculty of Science. “It is a fact that living beings glow. It’s a very weak glow, but it’s there and visible with very sensitive cameras.” According to the study, the light involved is extremely weak, ranging from 10 to 1,000 photons per square centimetre per second across a spectral range of 200 to 1,000 nanometres. For comparison, a nanometre is one-billionth of a metre and is commonly used to measure wavelengths of light. Detecting emissions at such low levels requires highly specialized equipment. To study the phenomenon, researchers used electron-multiplying charge-coupled device (EMCCD) and charge-coupled device (CCD) cameras. These imaging systems are designed to detect extremely small amounts of light, including individual photons, while minimizing background noise. The technology allowed researchers to capture signals that would otherwise be impossible to observe. The team worked with the Human Health Therapeutics Research Centre at the National Research Council of Canada (NRC) in Ottawa to examine photon emissions in mice. Researchers took two-hour exposure images of the animals before and after death and compared the results. “We saw that the level of light that they emit – this biophoton glow – is distinctly different between living and dead animals,” says Dr. Daniel Oblak, PhD, an associate professor in Physics and Astronomy and the corresponding author of the study. The images showed a clear decrease in photon emissions after death across the entire body of each mouse. According to the researchers, this provided direct evidence that living and dead tissue produce different levels of ultraweak photon emission. “It’s a very small amount and it’s, of course, very tricky to detect,” Oblak says. The study grew out of discussions between Simon, whose research interests include quantum biology, and Oblak, whose work focuses on detecting light for quantum communication experiments. Quantum biology is a field that explores whether processes described by quantum physics, which studies matter and energy at very small scales, may also play a role in living systems. “Since I work as a quantum physicist on light detection for quantum communication, I thought that experimentally we have a lot of the tools to be able to detect the light,” Oblak explains. The researchers also investigated UPE in plants and found that the light changed in response to stress. When plants were exposed to higher temperatures or physically injured, their photon emissions increased. Chemical treatments also affected the glow. Among the substances tested, the local anesthetic benzocaine produced the strongest emission response when applied to injured plant tissue. These findings suggest that ultraweak photon emission is closely linked to biochemical and metabolic activity inside living organisms. Metabolism refers to the chemical reactions that allow cells and organisms to stay alive and function. Because these reactions change when an organism experiences stress, injury or disease, researchers believe UPE may provide a way to monitor those changes. The researchers stress that the glow is a physical and biological phenomenon, not a metaphysical one. Oblak says more research is needed to understand exactly how the light is produced and what information it may reveal about the condition of living tissue. “We must understand what that is to figure out what’s happening,” he says. “If we can understand how that relates to certain influences on the body – stress, diseases – then that could be used as a diagnostic tool.” The researchers believe the technique could eventually help scientists study health and disease without invasive procedures. Because UPE can be measured without adding dyes, markers or labels, it may offer a way to monitor whether tissue is healthy, damaged or alive. In plants, it could help researchers better understand how organisms respond to injury, heat and other forms of stress. While the work is still in its early stages, the study demonstrates that ultraweak photon emission imaging can provide a non-invasive and label-free way to observe biological activity. Researchers say the approach could become a useful tool for studying vitality, stress responses and other important processes in both animals and plants. Source: University of Calgary, ACS publication This article was generated with some help from AI and reviewed by an editor. Under Section 107 of the Copyright Act 1976, this material is used for the purpose of news reporting. Fair use is a use permitted by copyright statute that might otherwise be infringing.
  • Recent Achievements

    • One Year In
      hhgygy earned a badge
      One Year In
    • One Month Later
      AMV earned a badge
      One Month Later
    • Week One Done
      AMV earned a badge
      Week One Done
    • Collaborator
      ryansurfer98 went up a rank
      Collaborator
    • One Month Later
      Eurosoft10 earned a badge
      One Month Later
  • Popular Contributors

    1. 1
      +primortal
      513
    2. 2
      +Edouard
      171
    3. 3
      PsYcHoKiLLa
      82
    4. 4
      Steven P.
      74
    5. 5
      Michael Scrip
      72
  • Tell a friend

    Love Neowin? Tell a friend!