PDA

View Full Version : Operating System Development



Tux.Ice
December 31st, 2007, 05:33 PM
Hello,

Me and a few buddies have decided to develop an operating system much like linux.

i am in need of graphic designers (i am one myself), and programmers(c, c++, python)

The operating system will be able to open .exe, .deb , .rpm, .tar.gz, as well as the other standard formats.

to ask specific questions please go to glacier.techotec.com/forum

tux.ice

Ozor Mox
December 31st, 2007, 06:12 PM
Wow. Er, do you mean an operating system kernel like Linux, or a complete operating system like GNU/Linux distributions?

Have you considered contributing to one of the various open source operating system projects already out there?

fatality_uk
December 31st, 2007, 06:15 PM
http://mikeos.berlios.de/

This is quite cute

EdThaSlayer
December 31st, 2007, 07:10 PM
That is quite a daunting task you have set there sir. GNU/Linux can do all the things you are proposing but that has been in development since the 1990's.

science4sail
December 31st, 2007, 07:16 PM
well, one could combine several projects...

namely:
ReactOS
Wine

for the exe part, but that's just a kludge...

LaRoza
December 31st, 2007, 07:17 PM
well, one could combine several projects...

namely:
ReactOS
Wine


ReactOS isn't usable yet.

EdThaSlayer
December 31st, 2007, 07:42 PM
ReactOS isn't usable yet.

It really needs a lot of work, I tried a live-cd of React OS, and it has quite a few bugs(and needs to look a bit prettier). :)

M$LOL
December 31st, 2007, 07:45 PM
I wouldn't be too worried about it being ugly, functionality is what I would value. But yes, it's nowhere near usable yet, although the source code is fun to play with :)

Lster
December 31st, 2007, 07:50 PM
May I say as someone who was and still is developing a kernel, you're in for a lot of work! Good luck all the same!

Methuselah
December 31st, 2007, 07:59 PM
OS development is interesting and easier now with the greater availability of VMs.
Something more than a toy will probably be hard to achieve though.
Getting devices to work will be especially tough.

I have mixed feeling about ReactOS.
Everytime I think of helping, I wonder if my efforts wouldn't be better spent somewhere else. The end result would be liberating (ReactOS in a VM could replace windows for most uses especially if 3D acceleration through VMs becomes possible) but it seems like it would be better for it not to be necessary to emulate windows.

Ah well, I guess reality is what it is.

az
December 31st, 2007, 08:01 PM
Hello,

Me and a few buddies have decided to develop an operating system much like linux.

i am in need of graphic designers (i am one myself), and programmers(c, c++, python)

The operating system will be able to open .exe, .deb , .rpm, .tar.gz, as well as the other standard formats.

to ask specific questions please go to glacier.techotec.com/forum

tux.ice

If you are starting from the point of wanting it to look a certain way and be able to open standard formats, I would tend to think you have no idea how much work is ahead of you.

You don't build any program from the ground-up by trying to make it look a certain way. That's the last step.

It sounds to me as if your needs would be served a lot better by taking an existing OS and theming it.

insane_alien
December 31st, 2007, 08:29 PM
well, one could combine several projects...

namely:
ReactOS
Wine

for the exe part, but that's just a kludge...

those projects are pretty heavily intertwined.ReactOS is (at the moment) just standalone WINE with some bits stuck on.

forrestcupp
December 31st, 2007, 08:34 PM
If you are starting from the point of wanting it to look a certain way and be able to open standard formats, I would tend to think you have no idea how much work is ahead of you.

You don't build any program from the ground-up by trying to make it look a certain way. That's the last step.

It sounds to me as if your needs would be served a lot better by taking an existing OS and theming it.

That's true. You're not going to need a graphic designer for a long, long time.

Nano Geek
December 31st, 2007, 08:40 PM
If you are starting from the point of wanting it to look a certain way and be able to open standard formats, I would tend to think you have no idea how much work is ahead of you.

You don't build any program from the ground-up by trying to make it look a certain way. That's the last step.

It sounds to me as if your needs would be served a lot better by taking an existing OS and theming it.


That's true. You're not going to need a graphic designer for a long, long time.I'd have to agree with both of these statements.

Developing a new OS is huge. It will take you years of work just to make it somewhat usable.

Make sure you're ready to do that much work before you start.

Tux.Ice
January 1st, 2008, 04:49 AM
it will be like gnu/linux but yes we will be developing our own kernel by changing linux source code.

if anyone wants to contribute graphics or even coding please do

Tux.Ice
January 1st, 2008, 04:50 AM
i do know that graphics and gui is last step, kernel being first , then a terminal like environment then gui

smartboyathome
January 1st, 2008, 04:59 AM
I would say if you are going to change the Linux kernel source code, maybe develop a better software installer into the kernel (no ./configure, make, make install, just one command). Also, it would be very hard to interwind all those install formats since they install to different places normally (for example, .exe installs to C:\ while deb/tar/rpm/etc installs to /. I would suggest looking at the gobolinux project for an idea if you want to change the filesystem to accomodate this.

kopinux
January 1st, 2008, 05:13 AM
ahhh! the spring time of youth!

until they learn it the hard way. it will take years to have a decent OS, maybe 10 years or so, and even with the stash cash of billions of $ like microsoft, they cant even buy time.

the_darkside_986
January 1st, 2008, 05:23 AM
Yeah really... the linux kernel is heavily developed, and distributions are heavily polished, but half of the distros don't even recognize my soundcard.

Why not just make higher level software that can "open" or "read" those formats? You know, you can do basic binary file input/output without having to bypass the OS or make a new OS.

If you mean opening and using the files as they are intended, linux can already be customized to do that. I've heard of people installing the redhat package manager in Debian or Ubuntu. I'm sure it is possible to associate the "exe" file type to be opened with "wine" in the nautilus settings. (Well, I'm not sure if it can distinguish between .NET executables and Win32 programs. By the way, does anyone know of a quick and safe way I can make my mono programs have a .mono extension instead of .exe? It seems so icky having exe's in Ubuntu...)

macogw
January 1st, 2008, 06:12 AM
ReactOS isn't usable yet.

My friend Shane says it can play a bunch of his Windows games that don't work with Wine.

M$LOL
January 1st, 2008, 03:18 PM
Yeah, a lot of people have good experiences with it from what I've heard, but it's far from finished. Certainly in my experience it's nowhere near useable, it won't boot natively on any of my machines, and it freezes randomly in my VM. There's still a lot of work to be done on it before it can even be considered for widespread use.

Tux.Ice
January 1st, 2008, 03:42 PM
i was thinking of using a script so that when an application prompts to install to c:\ refine the directory to \ would this work??

Tux.Ice
January 1st, 2008, 10:21 PM
would anybody like to help??

Tux.Ice
January 1st, 2008, 10:40 PM
**bump**

Namtabmai
January 1st, 2008, 11:52 PM
would anybody like to help??

Errr... perhaps you should review the posts in this thread. Creating a new operating system is no small task. Not to mention a few of your posts in this thread show you naivety in the subject, for example


i was thinking of using a script so that when an application prompts to install to c:\ refine the directory to \ would this work??
Suggesting that an operating system should, perhaps blindly, install into the root of a hard disk is just asking for trouble.

If you really want to go ahead with this project, I suggest doing the most important thing your project needs first. Planning.
Plan out, what you need to do, how you're going to do it, a time line of the events you need to achieve along with which parts of the project need to be completed before moving on to the next stage.
If you're starting from the ground up, get your kernel up and running and show people the benefits over Linux. This alone will probably take you a couple of years. Once you have a stable base such as this which can be shown to have major benefits over the Linux kernel then you'll find that you have lots of people wanting to help out, especially when you've got your project plan sorted and people can start asking to be involved with the various aspects of the OS that you plan on creating.

Tux.Ice
January 2nd, 2008, 12:17 AM
ok i shall do this - ty

Tux.Ice
January 2nd, 2008, 12:22 AM
any body have an idea where i can download the kernel source

Ebuntor
January 2nd, 2008, 12:59 AM
any body have an idea where i can download the kernel source

Here you go: http://www.kernel.org/

I wish you good luck with your project, as people stated in previous posts it will take many many years before you even have a basic OS.

You'd better review everyone's replies because I doubt you've got a clear picture of what you're getting yourself into, it's a lot of work, seriously.

zipperback
January 2nd, 2008, 01:08 AM
would anybody like to help??


Rather than try and get people from the Ubuntu community to try and build a new operating system from the ground up, why not try and improve on the very powerful and open source operating system that we already have.

And just so you know, I do have a great deal of experience with computers and programming. I've been using computers since the mid 1970's.

- zipperback
:popcorn:

bruce89
January 2nd, 2008, 01:29 AM
It took me about 5 hours to write a GTK+ widget, goodness knows how long it would take someone to write a kernel.

Linuxratty
January 2nd, 2008, 01:33 AM
why not try and improve on the very powerful and open source operating system that we already have.

- zipperback
:popcorn:

I'm in agreement with the other folks here...
Why reinvent the wheel?
There are plenty of Linux distros you could work with who could benefit from your services...Pick one..
Or if you want to work with React, or BSD, go for it.
I really think you should work on what is already out there,rather than starting from scratch.
Just my two cents worth.

p_quarles
January 2nd, 2008, 01:34 AM
It took me about 5 hours to write a GTK+ widget, goodness knows how long it would take someone to write a kernel.
I think it took Torvalds about a year to write the original (0.1) kernel -- but A) he's not a normal person; and B) it had nothing even remotely approaching the functionality of the current kernel, which is a massive, collaborative project that's been going on for nearly two decades.

And let's not even mention Hurd. ;)

samwyse
January 2nd, 2008, 01:35 AM
any body have an idea where i can download the kernel source

Sorry, but I don't think this project is going anywhere.

bruce89
January 2nd, 2008, 01:40 AM
I think it took Torvalds about a year to write the original (0.1) kernel -- but A) he's not a normal person; and B) it had nothing even remotely approaching the functionality of the current kernel, which is a massive, collaborative project that's been going on for nearly two decades.

And let's not even mention Hurd. ;)

I forgot about Hurd, now I feel better.


Sorry, but I don't think this project is going anywhere.

Linux? It certainly is.

If you meant this new one, fair enough.

Tux.Ice
January 4th, 2008, 03:46 PM
yes yes i know i have now decided to just modify the kernel slightly and basically build a linux distro on top of it - and as long as i create a credits sheet i dont even have to call it glacier linux. :) ty for your criticism

bobbocanfly
January 4th, 2008, 04:00 PM
If you are really serious about writing operating systems you could try hacking Minix's code. It is used in university Operating Systems courses and was designed to be educational. There are some really simple but satisfying hacks that can be done by changing one string in one functions, like editing the announce() function which displays the "Minix Version x.x.x - Copyright XXXXX". Thats obviously to get started with compiling it etc. there is much more stuff you can do with it.

smartboyathome
January 4th, 2008, 05:45 PM
If you do want to run Windows programs on it, you will HAVE to use something like gobolinux's gobohide in order to hide the symlinks for the folders.

mips
January 4th, 2008, 08:33 PM
Maybe try the programming forum?

hockey97
January 4th, 2008, 09:33 PM
since we are on the topic OS development.

I want to ask you guy's what a Kernal is really??
and what is really an .iso image?

DUDE_2000
January 4th, 2008, 09:54 PM
the one thing that comes to mind here, is trying to merge the reactos kernel with the linux kernel, then mess around with parts of debian and fedora, throw a DE on top of that, and start testing.

bobbocanfly
January 4th, 2008, 10:37 PM
since we are on the topic OS development.

I want to ask you guy's what a Kernal is really??
and what is really an .iso image?

The Kernel is the low-level part of an operating system that connects between the CPU, RAM and other devices and allows applications to run. http://en.wikipedia.org/wiki/Kernel_%28computer_science%29 will help a lot!

An iso image is a raw copy of an ISO9660 (CD) Filesystem. Its basically just the whole of a CD copied onto a comuter.

forrestcupp
January 5th, 2008, 01:18 AM
Maybe it would be better for you to use the Linux Kernel, and just program a new window manager or desktop environment.

Or if you really want to get compatibility with .exe's why not just help with the Wine project.

You don't really need to start from the ground level. Why not just create a DE, then create a new distro and figure out a way to integrate Wine into it seamlessly, then really go to work helping Wine become more compatible with Windows?

This would still be a huge project, but nothing like creating a kernel and doing all of the stuff I mentioned. The Linux Kernel has taken 17 years to get where it is now, with many contributors.

Kernel Sanders
January 5th, 2008, 01:42 AM
Personally, I suggest making contributing to PC-BSD. That project would be awesome if it had more support.

mehaga
January 5th, 2008, 01:50 AM
Reading this thread, I keep asking myself how did the OP, who:

- started developing a new OS, but decided to plan it out only after someone here suggested he should do so,
- wanted to develop it by modifying linux kernel without knowing where to find it's source code
- ...

get so many serious answers? :confused: :)

Tux.Ice
January 6th, 2008, 09:56 PM
actually i knew about kernel.org i was just wondering if there was another site.

Namtabmai
January 6th, 2008, 11:09 PM
actually i knew about kernel.org i was just wondering if there was another site.

Ah so you want the site which has the Python implementation of the Linux kernel for easier hacking? You want, http://www.pythonkernel.org/.












Now be honest please, who actually clicked the link? :)

science4sail
January 9th, 2008, 12:23 AM
Ah so you want the site which has the Python implementation of the Linux kernel for easier hacking? You want, http://www.pythonkernel.org/.


:lolflag:

Tux.Ice
January 10th, 2008, 12:48 PM
hahaha :lolflag: