This topic comes up every few months. The reason why creating such a GameOS is not a particularly good idea is that such an OS would be useless without good & widespread driver support, and the companies that make hardware (videocards, soundcards, etc) aren''t going to waste time porting drivers to your OS. And in most cases (these days, things were different a few years ago), they don''t even publish the specs that would allow you to write your own.

Really doing something like this to the point where it would be useful would be a ton of work and you''d always be playing catch-up because of the driver issues. Look at Linux and even BeOS. They were pretty well known and still have plenty of driver issues themselves, and Linux, for one, has been in development for about 10 years now.

If you want to do something like this for the sake of learning, go ahead, its probably a nifty way to learn low-level coding. But its totally impractical other than for educational purposes. That''s not a flame, just the truth.

A game OS would be so cool, you could do stuff like prevent stupid background and taskbar-like programs from running and taking up resources/memory. but drivers, I agree would be a real problem, unless if you could somehow clone (reverse engineer?) the way Windows uses DirectX drivers, like mr_jrt said but in different words.

Anyway, I wanted to ask someone reading this thread about what documentation or books I should read if I wanted to make an OS. I imagine there would be a lot of details. I would really like to know stuff like how to make a hd partition bootable in assembler, etc.

Directed at anyone out there....if you want info on any of the stuff I''ve learned (boot sectors mainly and some basic OS related things), give me a bell. My details are in my GDNet profile.

Please though, nothing too involved, I''m still learning myself!


Waassaap!!

i think a gaming os would be great. it could be sort of like windows but only for games. That way all resources could be decaded to the game. hell, we all know microsoft has the resourcs to do it. and everyone could just run a dual boot or whatever. but of course that would never work since most people don''t even have a clue about dual booting and stuff and people would rather go buy a game console. it would be nice in a perfect world, but this isn''t a perfect world.

The idea sounds stupid and is impractical . It doesnt address any OS specific problem affecting programmers or gamers .

quote:

---

I think a gaming os would be great. it could be sort of like windows but only for games.

---

You sound very confused . HAve you heard of the X-Box ? It runs a windows OS , even uses DX9 , yet its specifically for games , is this whaty you are looking for?

quote:

---

A game OS would be so cool, you could do stuff like prevent stupid background and taskbar-like programs from running and taking up resources/memory

---

Get a console then . The reason why those program run is because people use their computers for other things other than games.Besides any self respecting gamer would know how to turn off unnecessary programs.

quote:

---

For me, the perfect GameOS must be multiplatform (computers & console) and use only few standard API: GLUT + OpenGL + POSIX + another for sound

---

Sounds to me like you are looking for a multiplatform all inclussive SDK , and before I forget A CONSOLE IS A COMPUTER.

quote:

---

It must be very simple and little with no 3Go of stupid DLLs.

---

I dont understand what you mean , maybe Im getting old.

quote:

---

I personally like the idea, I think that the gaming os should be on a platform like the computer not a dedicated console, with this game os, it could be useful to all the other programs that ran on it.

---

Hugh wtf ? What is the industry definition of a platform ? Isnt a console a platform?what do you mean by

quote:

---

it could be useful to all the other programs that ran on it.

---

Maybe Im getting too old.

You guyz are not making any sense (sound like a bunch of newbies) . You are not looking at a way to improve either game developing or game playing experience (probably coz you dont fully comprehand the industry , or the programming aspect of it , or even how OSes work infact I doubt if any of you have ever installed an OS from scratch). You need to ask yourself a few basic questions first ,

1) What is an OS?
2) How does an OS work?
3) How can the current OSes be improved ?
4) What is the common problem with the current games?
5) Is this problem an operating system , hardware, or a programing issue?
6) Is there a way around it or does it warranty writing a complete os?
7) If it warrants a new os what is the target hardware?
8) How is the new OS going to address the "problem" ?
8) How is the new OS going to appeal to the average Joe,hardware manufacturers(for drivers) , game developers?

Look at other oses for example BEOS , perfect for games and Multimedia Aps . You guys sound like youve never installed an OS before, let alone design/code one.

Game OS , rotflmao , sounds like ,

Newbie1-"Lets design a TV that shows only cartoons because we are tired of switching channels or looking at the TV guide to find the cartoon channels ."

Newbie 2-" It would so cool if we could make a TV for cartoons only.Yea we dont want to make a new channel that shows only cartoons ."

NewBie3-"Yea dude thats an awesome idea , we want to make the whole tv that shows cartoons only because it would be so cool "

NewBie4-"Yea , and it wont show any of the current cartoon channels , because the whole tv is gonna be nothing but cartoons , wed call it cartoon TV"

NewBie5-"Yea dude , the perfect cartoon TV would be cross platform , we should be able to take the stuff out of a computer monitor and make it run our cartoon TV "



I hope the moderator has a good sense of humor and will keep this post .

Most people get thrilled by the thought of "coding an OS". They think it''s hard and they think it''s cool. It''s not. Basically, what you need to know is how the hardware works. I''m not talking about physics, electronics or math. I''m talking about interfaces. It''s like memorizing all the numbers in a phone book. I may be fun the first time you write your bootloader, but there are lots of boring stuff that needs to be done to get your OS up and running. And no matter how good it is there''s no chance that any company will ever write drivers if it hasn''t got any users. And who wants to use an OS that has no drivers?

And remember: Even though some games that currently doesn''t run very well on your computers might run better with your "gaming OS" it would take so long to make that if you got a job instead of coding it, then by the time you would have finished it (which by the way is very unlikely that you ever would, not because you''re not good enough, but because your team would probably lose interest in the project) you had already earned more than enough money to buy yourself a new computer with memory enough to run both the latest games, Windows and hundreds of useless memory-consuming tray applications.

Actually, I agree with Hoodrat on the issue..

The issue really isn''t windows.. Considering how it''s designed (with a microkernel, along with a HUGE ammount of dynamic libraries (DLLs) and other sort of drivers and handlers) is about as flexible you can get an OS.. All things like updated drivers (and/or DirectX) work BECAUSE it is like that.

The problems of today with PC games come from three main issues:
#1, Bugs - because of insufficient pre-release testing
#2, Lack of optimization - hardly anyone optimizes that much anymore
#3, The PC is pretty much THE most resource-eating bottleneck-packed personal computer that exist today!

I''d think that before we can talk of a gaming OS, we''d need a new gaming ''puter.

"Goodness reflects the light; and evil, bears the seed of all darkness. These are the mirrors of the soul, the reflections of the mind. -Choose well..." (Unknown)

How about, instead of getting a new computer, we get new programmers? I personally work on something just barely long enough to get it working before I start optimizing. As Kompi said, hardly anyone optimizes anymore; this is why a lot of games have these extremely high-end requirements; the programmers never took the time to get them working faster.

Also as Kompi said, and as every person on the planet with a PC knows, bugs abound. Probably about half of the people who post here are old enough to clearly remember the days when, if you had even a single bug in your program, it went straight back to your programming tool of choice. I personally know of games (and still play them on an Atari ST emulator ) that have never frozen, crashed, or had any other type of error while I was playing them.

Perhaps, when it comes to programming, we need to work on the quality of our programming. All of us. I don't preach to be perfect (and I go to a lot of trouble to make sure of that ), and I can only hope that someday we all will.

Edit - My grammar horrible was Fixed it, now...

Edited by - Wyrframe on July 3, 2001 6:57:04 PM

I agree completely with Wyrframe. To give an example, I recently wrote a basic set of DirectDraw wrappers that were meant to be a stable and solid foundation for graphics rendering in my engine. I started off by basing a lot of the code on the classes that came with the DirectX8 SDK, although with a tad different class design (I hate the MS way of thinking when it comes to OOP). In the MS code alone I found at least 10 bugs, ranging from stuff not working at all to rather embarrasing memory leaks and redundant code. Ended up almost completely rewriting everything in the end. However my code was far from perfect as well, yet I was under the assumption it was fine (since it ran and everything seemed well). Thankfully, I was also trying out my error logging class (that I spent a ridiculous amount of time on to get it to have some spiffy features), and I caught a truckload of bugs in my code, all by looking through this log file in fully verbose logging mode and finding lots of weird stuff going on, which allowed me to not only fix quite a few memory leaks, but also improve speed a noticable amount.

Now this is just simply taking the time to writing code that works. No real optomizing or anything, just straight proper interfacing with DirectDraw. Yet even that can be mucked up. Too many games simply dont have all their code tested properly while it is being written, and the result is a mess of hard-to-trace problems and huge speed loss

Unfortunately, they are able to get away with it.

quote:

---

Original post by Wyrframe
How about, instead of getting a new computer, we get new programmers? I personally work on something just barely long enough to get it working before I start optimizing. As Kompi said, hardly anyone optimizes anymore; this is why a lot of games have these extremely high-end requirements; the programmers never took the time to get them working faster.

---

A lot of younger people are entering the industry these days and they never had to deal with such issues as having to fit your entire game into 64KB. They have been spoilt by multi-megabytes of RAM and HDD space that they see as a space waiting to be filled.

In my current project, I decided to go through the menu system and do some resource optimization. I reduced 30 separate textures to just 5 with a bit of smart thinking and re-use of existing assets.

It is an art-form that is withering away. Systems like the Gameboy and Gameboy Advance are keeping it alive though. These systems reek of old-school game development, and that is a good thing.

quote:

---

Also as Kompi said, and as every person on the planet with a PC knows, bugs abound. Probably about half of the people who post here are old enough to clearly remember the days when, if you had even a single bug in your program, it went straight back to your programming tool of choice. I personally know of games (and still play them on an Atari ST emulator ) that have never frozen, crashed, or had any other type of error while I was playing them.

---

One major difference between the past and the present is that in those days, your game had complete control over the entire system. These days, the robustness of your game relies on the stability of many layers of software out of your control, including but not limited to Windows, device drivers, background processes, services, and any other applications that you have running such as ICQ or virus detection software.

quote:

---

Perhaps, when it comes to programming, we need to work on the quality of our programming. All of us. I don''t preach to be perfect (and I go to a lot of trouble to make sure of that ), and I can only hope that someday we all will.

---

It''s a small quest that I have been quietly working towards not only for myself, but all programmers that I work with.

Steve ''Sly'' Williams  Code Monkey  Krome Studios