A Collaborative Free and Open-Source OS?

Started by
115 comments, last by Tutorial Doctor 10 years ago
To me, the perfect operating system is one that is fast, user friendly, customizable/tweakable, and FREE!

I like IOS because of its speed and user friendliness, and I like Android because it is free and tweakable.

Of course, there is Linux, and most people would use Linux as a base for a custom OS (and it seems I would prefer one use Unix), but for now the best alternative we have/had is/was Ubuntu (they lost me after 10.10)

I have been dreaming of such an OS for a while, one that is a strong focus on all of the above. I think that all of the Linux folks would do better collaborating on another OS altogether rather than trying to "fix" linux.

Linux serves its purposes, but the reason Microsoft hit it big was because of the relative user friendliness of Windows. Apple hit it bigger because it had user friendliness and speed.

I should note that of course we don't want bugs (blue screens of death and such).

Apple can be beat, and if only I was as astute as some people, I would know just how to beat them.

Good idea? Bad Idea? Yeah, right?

They call me the Tutorial Doctor.

Advertisement
One problem with starting from scratch is that momentum is really working against you. Take, for example, Linux. When it started, computers were mostly single-core, there wasn't nearly the proliferation of device types that you see today, and the basic computing model of Unix was already well-established as a template to work from. In effect, it was much easier to write a small kernel and work from there. But architectures are quite a bit more complex now, with a lot more types of devices that need to be supported in order for your OS to be of any use. As architecture complexity has grown, the complexity of the Linux kernel has grown with it, along with the requisite engineering and bug-fixing required to make it work correctly. You could, of course, poach device models from Linux, and look to Linux for some inspiration on how to structure your kernel, but at that point you'd essentially just be duplicating Linux, so what's the point?

Also, I think you misunderstand what Linux is. Linux is the kernel, the bit of code that sits at the bottom running the show. Linux is not the Gnome desktop, nor is it the KDE desktop, nor Enlightenment, nor any other. They run atop Linux, sure, but they are not Linux. I suspect that when you refer to 'fixing Linux' you are actually referring to fixing the windowing and desktop environments that commonly sit on top of it, an assumption backed up by your subsequent comments about user friendliness. This really has very little to do, at all, with Linux itself and there is really not much need to completely redo everything from the core up just to fix some user friendliness problems.

Currently, the XWindows system which runs atop the kernel and provides frame buffer access does kind of stand as a major obstacle in fixing some of the usability issues. By anyone's assessment, X is a mess. Additionally, the fragmentation of the desktop environment communities hinders forward progress in the user friendliness arena, as each package and each maintenance team have their own ideas. Distribution fragmentation further complicates it, because now you have even more teams with even more dissenting opinions on what constitutes 'user friendliness'.

These are issues that stem not from Linux, but rather from the open and collaborative approach of an open source project. And basically, what you are proposing is to merely increase the level of community fragmentation by putting forth yet another option, with all its warts and knots and bugs and tradeoffs, one that will be starting from zero without the benefit of decades of bug-fixing and the combined experience of hundreds of engineers, device driver developers, and industry-funded experts.

I'd say that if you have real, concrete ideas on how user-friendliness can be established, that you start by building and tweaking a custom Debian distribution. That way, you're not spending decades re-inventing the wheel. Debian is a common starting point for many distributions (Ubuntu included) and with a bit of research you could probably tweak things the way you like.

standards.png

The "problems" with Linux (if you want to call these problems at all) is because it's open source. We have soo many flavors of Linux and none of them actually hit the spot to be *that* popular. This is the direct effect of having it open source. People have opinions. Your opinion is different to that of X in Japan, and Y in Russia. When you, X and Y starting to collaborate on an open source project, you will have arguments and opinions on where to go, so eventually you, X, and Y each have your own version of the software. This is the nature of open source. The direction of open source product becomes a scatter shot, rather than focused.

Ubuntu comes really close. Android also comes close, but that's mainly because Google was the driving force behind it. Without the massive funding and support by Google, Android wouldn't be where it is today. Any product needs some direction to get somewhere. It requires some authorative figure to mandate the direction. Even so, there are still people who don't like where things going, and take the liberty to take it on their own.

A perfect OS for me is a 'smart', multi multipurpose, multi multitasking, not free, not open source, very customizable, voice/gesture controlled OS with a great UI, always 'learning' and is very user 'friendly' and definitely not related to apple or linux, but a teenie weenie bit of Windows (can't wait to start making my version of J.A.R.V.I.S.).
Windows is perfect but i still want mine.

UNREAL ENGINE 4:
Total LOC: ~3M Lines
Total Languages: ~32

--
GREAT QUOTES:
I can do ALL things through Christ - Jesus Christ
--
Logic will get you from A-Z, imagination gets you everywhere - Albert Einstein
--
The problems of the world cannot be solved by skeptics or cynics whose horizons are limited by the obvious realities. - John F. Kennedy

A perfect OS for me is a 'smart', multi multipurpose, multi multitasking, not free, not open source, very customizable, voice/gesture controlled OS with a great UI, always 'learning' and is very user 'friendly' and definitely not related to apple or linux, but a teenie weenie bit of Windows (can't wait to start making my version of J.A.R.V.I.S.).
Windows is perfect but i still want mine.

I must admit, I loled.

A perfect OS for me is a 'smart', multi multipurpose, multi multitasking, not free, not open source, very customizable, voice/gesture controlled OS with a great UI, always 'learning' and is very user 'friendly' and definitely not related to apple or linux, but a teenie weenie bit of Windows (can't wait to start making my version of J.A.R.V.I.S.).
Windows is perfect but i still want mine.


I must admit, I loled.
It is very lolable.

UNREAL ENGINE 4:
Total LOC: ~3M Lines
Total Languages: ~32

--
GREAT QUOTES:
I can do ALL things through Christ - Jesus Christ
--
Logic will get you from A-Z, imagination gets you everywhere - Albert Einstein
--
The problems of the world cannot be solved by skeptics or cynics whose horizons are limited by the obvious realities. - John F. Kennedy

If an operating system boots fast, and runs all the software I need it to, that is all I desire from it. For average day-to-day use Mac, Windows, and most Linux distros all have a very similar work flow, as most applications run on all three.

Stay gold, Pony Boy.


I have been dreaming of such an OS for a while, one that is a strong focus on all of the above. I think that all of the Linux folks would do better collaborating on another OS altogether rather than trying to "fix" linux.

I don't think anyone tries to "fix" linux. Thats just how linux is designed to work.

Linux itself is just the kernel and drivers, the rest is general open source software, and anyone can create their own bundles.

All the distributions are just that, bundles of separately developed software shipped together with a linux kernel.

And since anyone can, and is encourage to do it, you get a lot of variants. Some call it confusing, others call it choice :)


I don't think anyone tries to "fix" linux. Thats just how linux is designed to work.

Exactly. At this point I can't really elaborate on the original intent of the design of Linux (whether or not it was ever meant to be able to be an OS for the general public).

I do think when a person makes an OS they target a specific type of use. I hear Linux is better for servers, but not for the desktop experience (at least its intent). Ubuntu passes, and some other distributions can also, but the architecture of Linux may not yield desirable results for the masses.

Or perhaps it isn't even the architecture of Linux?

I just got the impression people were trying to make Linux do what it wasn't designed to do just because it is the only free and reasonable base alternative OS to Windows.

Why not Unix?

If only making an OS was like making a peanut butter and jelly sandwich...

They call me the Tutorial Doctor.


I don't think anyone tries to "fix" linux. Thats just how linux is designed to work.

Exactly. At this point I can't really elaborate on the original intent of the design of Linux (whether or not it was ever meant to be able to be an OS for the general public).

I do think when a person makes an OS they target a specific type of use. I hear Linux is better for servers, but not for the desktop experience (at least its intent). Ubuntu passes, and some other distributions can also, but the architecture of Linux may not yield desirable results for the masses.

Or perhaps it isn't even the architecture of Linux?

I just got the impression people were trying to make Linux do what it wasn't designed to do just because it is the only free and reasonable base alternative OS to Windows.

Why not Unix?

If only making an OS was like making a peanut butter and jelly sandwich...

Again, Linux is just the kernel. Almost everything you as a user interact with on a daily basis is provided by a desktop environment running on top of that kernel, and all the workflow issues you might get are not due to the kernel which doesn't even know or care what you're doing at a high level, the kernel just creates a process, opens a file, discovers hardware devices, etc... none of this being immediately usable by a computer user, who needs software to interact with the kernel and interpret what it gets. What this means is that Linux isn't "designed" to do anything in particular, it's a general purpose kernel which has been used to great success in many different environments ranging from the average PC to supercomputers to microcontrollers and other embedded devices like routers. It provides an abstraction to the hardware, how this is presented to the user is completely out of its scope.

The Linux kernel isn't user-friendly. It isn't user-unfriendly either. In fact, it's not used by the user at all! That's why it's called a kernel, you don't see it because it's hidden away under the protective shell of userland software. That software, which the user interacts with, depends on what the user wants to do with his computer, and therefore these are packaged in distributions for ease of, well, distribution, some are targeted at specific user groups (niche distributions with very specific software installed), others are for power users, and yet others are for the general public and have similar interfaces to popular operating systems like Windows. You can roll your own according to your own specific needs (in fact, this has probably already happened - you started with some distribution and then tweaked it over time to customize it to your needs), though unless it's usable by lots of different people there's probably no point in sharing it.

Either way, I don't believe in strict standardization, the past has shown that humans are pretty crap at deciding on standards and tend to make them either too vague (making them frustrating or downright impossible to work with) or too strict (physically preventing new ideas from adhering to the standard). Rather I believe in interoperability, where things might not be exactly the same, but behave (or can be made to behave) in similar if not identical ways. That is a much easier and realistic goal than ensuring everyone always uses 100% the same hardware or software, and has been shown to work well in general (see the hardware abstraction layer, cross-platform software, and so on).

“If I understand the standard right it is legal and safe to do this but the resulting value could be anything.”

This topic is closed to new replies.

Advertisement