OSDev.org

Hi,

At the beginning i would like, as i am new here, to greet everybody and sorry for my poor English and potential mistakes. I hope this will not cause any misunderstands.

My name is Rafal and i am a member and one of Arcon Group! founders. I write here because we are looking for more developers. I hope our project development state will not discourage You - it is in planning and early pre-alpha state.

At first, i will write something about our team, if You let me to... We originally started development as far back as 1999 (as BetaComp Team). Although the design mock-ups are much different to today's result, the goal has remained the same. Arcon Group! is the result of many peoples work. Our team is an assocation of individuals who have made common cause to create some Free Software for everyone, which members are from all over the world. We want to help both - advanced users and those who aren't computer experts. Our goal is to solve each, even hardest reported problem and we support Free Software products (actually we support Gentoo Linux providing its users portage overlay: You can read more about it here: http://arcong.ath.cx/index.php?option=c ... &Itemid=30)
More information about our group can be found at our web site: http://arcong.ath.cx/

But let's see the hearth of the matter - Cinay. It's completely new, multimedia operating system and environment originally based on Linux. It branched in 2007 in order to develop a radically different approach to concurrency, SMP, and most other kernel subsystems. It belongs to the same class of operating system as Linux, BSD and is based on the same UNIX ideals and APIs. It main goal is to provide a desktop operating system that's fast, stable and easy to use. We want to build an operating system that meets most requirements desktop users have, like easy installation of software or sharing an internet connection.

Cinay is going to be a multi-year project. It will take a lot of groundwork to even approach the goals we outline here. By checking our various goal links, you can bring up position papers on various aspects of kernel and system design which the project hopes to accomplish. First and foremost among all of our goals is a desire to be able to implement them in small bite-sized chunks, while at the same time maintaining good stability for the system as a whole. While the goals are not listed in any particular order, there is a natural order to things which should allow us to advance piecemeal without compromising the stability of the system as a whole.

One of our goal is to create a flexible dual-purpose caching infrastructure which mimics the well known compressed caching model over a broad range of configurations. It is the introduction of a new level into the virtual memory hierarchy. Specifically, RAM is used to store both an uncompressed cache of pages in their 'natural' encoding, and a compressed cache of pages in some compressed format. By using RAM to store some number of compressed pages, the effective size of RAM is increased, and so the number of page faults that must be serviced by very slow hard disks is decreased. Our aim is to improve system performance.

CinayOS ups the power of 64-bit computing! Build and run a new generation of 64-bit applications that address massive amounts of memory. It takes 64-bit computing to the next level, while maintaining full performance and compatibility for any existing 32-bit applications. It delivers 64-bit power in one, universal OS. Now alls application frameworks, as well as graphics, scripting, and the rest of the system are 64-bit. It delivers 64-bit power to both Intel- and AMD-based PCs, so you don’t have to install separate applications for different machines. There’s only one version of CinayOS, so you don’t need to maintain separate operating systems for different uses.

Cinay is based on a kernel modelled around Linux, which is one of the most reliable and robust operating systems. The development is highly focused on performance. The system manages resources efficiently and reacts quickly even when many processes are running simultaneously. Introduced optimalisations make it working on every CPU with at least SSE2. Completely rewritten libraries introduce ability to handle 128-bit computations in a single clock cycle and to use known CPU instructions-set like SSE, SSE2, MMX(+) (and SSE3, SSSE3, 3DNow(+) if they are available).

Cinay OS can run on any AMD64- or EM64T-based computer equipped with at least 512MB RAM and modern graphics card. As it's in pre-alpha and planning state, supported hardware components list is not available.

It will include a Core Animations Library! Creating dynamic scenes can get confusing without a systematic approach. OpenGL Accelerated desktop that seeks to provide a free, open source desktop experience to the community. With CinayOS the rather esoteric concept of the computer desktop is brought down to a more human level, allowing for a more native and intuitive understanding of your workspace. But it's not only a desktop! Using Core Animations, developers can create snazzy animations in their programs! It dynamically renders layers together, complete with transparency effects and image filter and effects. When content changes, updates it automatically.

More technical details can You find here: http://arcong.ath.cx/index.php?option=c ... &Itemid=50
and features list, here: http://arcong.ath.cx/index.php?option=c ... &Itemid=42

Please note that above description doesn't present real Cinay development state. As OS is in planning state, it is only suppose.


I am looking forward to hearing from You.
Thanks!

Hi, and welcome.

You may consider it a compliment that I'm responding to your message above others in the 'we want good developers' vein as you have actually put some thought into what you want to do.

I note, however, that most of your post relates to elements of an operating system that are separate from the kernel, e.g. a 'multimedia operating system', easy software installation, internet connection sharing and graphics libraries. There already exist packages for most Linux distributions which allow you to do all these things.

The most interesting part of you post, at least to me, relates to your compressed memory system. Am I right in thinking that you would use part of system memory as an extra layer between usable RAM and a swap file? If that is the case, then you must also realise that using part of your RAM as a compressed store will increase the number of page misses, as you have a reduced amount of free physical memory in which to put your actual data. Do you know of any articles discussing such a system (and particularly the performance gains from it)? I should be interested to read them.

Unfortunately, having browsed your subversion repository (and from your original post) I note that you have just copied the entire Linux 2.6 source tree. Modern Linux is notoriously difficult to understand. Do you have any experience with the kernel (as you would surely need to make such dramatic changes to the memory system)?

morhpeouss wrote:Completely rewritten libraries introduce ability to handle 128-bit computations in a single clock cycle

Can someone verify me on this: am I right in thinking there is no 128-bit operation that completes in a single clock cycle?

morpheouss wrote:Linux, which is one of the most reliable and robust operating systems

You might get some arguments about that here . Personally, I don't believe Linux is a well designed operating system, and it is now over 15 years old. Maybe its just I have something against monolithic kernels... Also, I don't like the GPL, and by basing your OS on the Linux kernel you require your OS to also be GPL (I guess that's my main problem with it).

Anyway, good luck with your project.

Regards,
John.

jnc100 wrote:Hi, and welcome.
The most interesting part of you post, at least to me, relates to your compressed memory system. Am I right in thinking that you would use part of system memory as an extra layer between usable RAM and a swap file? If that is the case, then you must also realise that using part of your RAM as a compressed store will increase the number of page misses, as you have a reduced amount of free physical memory in which to put your actual data. Do you know of any articles discussing such a system (and particularly the performance gains from it)? I should be interested to read them.

Compressed caching is the introduction of new layer in virtual memory hierarchy -- Compressed Cache. It compresses and stores pages that would otherwise have been swapped to slow disks or freed under memory pressure. This effectively increases RAM space and avoids /reduces accesses to slow disks. This basically takes advantage to rapidly increasing CPU power, faster, lower latency memories and sluggish hard-disk speed improvements. These all should work with kernel preemption.
Each open file has a separate radix tree to maintain its pages in page cache. So, in effect, each open file has its own page cache. The offset within the file is used as key to locate the corresponding page in memory.
All swap cache pages are part of a single swapper_space – a single radix tree maintains all pages in the swap cache.swp_entry_t is used as a key to locate the corresponding pages in memory.

jnc100 wrote: Unfortunately, having browsed your subversion repository (and from your original post) I note that you have just copied the entire Linux 2.6 source tree. Modern Linux is notoriously difficult to understand. Do you have any experience with the kernel (as you would surely need to make such dramatic changes to the memory system)?

You should see, we have just started some works with Cinay.
I guess , You think we are making another Linux distribution, which of course is not true. We have just kernel which is modelled around linux kernel, which should be light, but actually is not . This allowed us to save time. We will improve its performance and optimise it and whole system as i have already written. it will work only on modern computers.

jnc100 wrote:

morhpeouss wrote:Completely rewritten libraries introduce ability to handle 128-bit computations in a single clock cycle

Can someone verify me on this: am I right in thinking there is no 128-bit operation that completes in a single clock cycle?

Of course, there is. The first operating system which has this implemented is Mac OS X 10.5. It's possible thanks to enhanced 128-bit SSE3 vector engine improved in Core 2 Duo... That probably the reason why You haven't heard of that.

jnc100 wrote:

morpheouss wrote:Linux, which is one of the most reliable and robust operating systems

You might get some arguments about that here . Personally, I don't believe Linux is a well designed operating system, and it is now over 15 years old. Maybe its just I have something against monolithic kernels... Also, I don't like the GPL, and by basing your OS on the Linux kernel you require your OS to also be GPL (I guess that's my main problem with it).

Maybe not well, but without doubt better than in other OSes. And don't forget , we can improve it...

About license... Every kind of license has bot good and bad points... I personally think GPL is good license...