Writing a OS...

[MystikShadows]

From Operating System to battle of the minds, wow, this promises to be to some powerful OS ;-). lol

In the meantime, this is to the original poster (message list is so long I don't remember who it was ;-). Have a look here, you just might like it.

http://www.nondot.org/sabre/os/articles

As for the rest of you on this thread, how much traffic would be saved on the QBasicNews server is replies were aimed at answer the question at hand? ;-). Seriously, anyone wants to start an OS, fine by me, if I can answer a question, awesome :-). But looking at the so called "powerful" OS, I say there's plenty of work to do. Windows, Linux, and the likes aren't really powerful when you "really" think about it. There better than 15 years ago....but to me, they are heading in the wrong direction. Especially our beloved Windows. They are shooting themselves in the foot and pretty soon there wont be a foot to shoot at anymore ;-).

I'm no prophet, but if I was to make a prediction, new and innovative OS/File management Idea would be where I'd be throwing myself. Get cracking people :-).

[TheBigBasicQ]

....but to me, they are heading in the wrong direction. Especially our beloved Windows. They are shooting themselves in the foot and pretty soon there wont be a foot to shoot at anymore ;-)....

Please explain this part.

[jatos]

I think anyone who waste's time provoking petty arguments is compeletly childish.

Anyway as has already been said you need to know asm or c++ to write a good OS.. I would suggest you also learn the technical stuff like how binary and hex works... (might sound stupid but its important when your doing this kind of stuff)

[The mega-mighty SPP]

I would suggest you also learn the technical stuff like how binary and hex works... (might sound stupid but its important when your doing this kind of stuff)

:rotfl: Yeah, I would suggest you also learn even more technical stuff like how flow control and CPU flags work and that kind of stuff.

[BastetFurry]

Ok, before you all invent your new OS and waste your time there, why dont you go and help the guys over at reactos.com?

[Moneo]

I would suggest you also learn the technical stuff like how binary and hex works... (might sound stupid but its important when your doing this kind of stuff)

:rotfl: Yeah, I would suggest you also learn even more technical stuff like how flow control and CPU flags work and that kind of stuff.

Yeah, "that kind of stuff" includes:
* I/O handlers for every device your OS will support.
* Dynamic memory allocation.
* Task scheduling.
* Design of Task Control Blocks.
* Interrupt handlers for: the clock, and all I/O including comm ports.
* Primary and secondary interrupt handlers considerations.
* Multi-tasking considerations.
* Etc.

Designing and writing an OS is probably the most complicated endeavor that anyone (or a group) can ever get into. Unless, it's an OS for a particular, limited environment to run a single application --- like perhaps an OS for a microprocessor based onboard computer for an automobile.
*****

[Mac]

Well, Justin (QAZIOP.exe) never came back.

I wished he would clarify. It seems that the first program a newbie wants to write is "an OS". I keep wondering why.

Mac

[Moneo]

Maybe it's like the first car a kid wants --- a Formula 1 Ferrari.
*****

[relsoft]

Maybe it's like the first car a kid wants --- a Formula 1 Ferrari.
*****

And when you're old enough, you'd be happy with a Kia. :*)

[neuro]

OS can be pretty ambiguous. Do you mean just the kernel? Do you mean the kernel AND a bunch of hardware drivers? Would you include things like TCP or iptables to be part of the OS? And what about the shell or (G)UI?

Some people argue there isn't enough documentation. Hardy har har. Just go to www.kernel.org and look at the linux source code. Pretty much any specific question you have can be answered with google.

Many higher level CS college programs involve kernel programming at some point or another.

Other interesting topics for fundamental programming include:

writing a malloc() system
writing a compiler
writing a garbage collector
writing a thread emulator

It's not a waste of time to try out writing these things. It is however slightly a waste of time if you want to have a comprehensive system that is marketable. Every major OS out there has been contributed to by hundreds if not thousands of people. A better course of action would be to make a donation of code or time to one of the currently existing OSes.

But if you just want to get to understand how the computer works without making a bunch of fancy features, then yeah, give it a shot.

-- Eric / neuro[/list]