I mean, after having used Linux for 10 years or more, I am feeling that maybe at some point of time Linux/*nix has probably resulted in too little innovation in the Open Source operating system area and too much re-inventing the wheel has been the result of it all.

Take the whole Xorg/DEs/WMs as example. It has too many layers, too much configuration, too many choices. All just to get a working desktop you not only need Xorg and its dependencies but also any choice of DEs and WMs as layered on top of it. And each desktop has its own way of configuration that sometimes conflicts with the underlying Xorg configuration or other desktops. Though modern Linux distros solve the problem with automatic configuration and to some extent adhering to freedesktop standards, the underlying mess still exists. And yes, too much reinvention of the wheel because of the basic structure of *nix which allows for this flexibility.

Take sound and multimedia. Again, there are so many kinds of sound systems, it's a bit of a mess. OSS, ALSA, pulse audio, then each desktop has its own sound-systems which work as some kind of middleware, like Phonon/KDE, gstreamer, etc. etc. Too much work just for a working sound setup.

Another example: an ancient UNIX style file permissions system which probably could have been much more enhanced to keep up with modern requirements if people were willing to break the historical baggage.

I recently tested out Haiku OS (still in Alpha) and it's like a breath of fresh air though it's clearly based on BeOS - a proprietary OS from the late 90s. But it's a tightly integrated little desktop OS with a clear focus and targetted at media and home users. And it has a simple clear single user interface.

I sometimes long for a simple non-Windows, non-*nix desktop operating system that just works as a desktop or whatever it's intended for and doesn't have too much of choice and follows a KISS philosophy. Haiku seems promising but there are too few users and developers out there at the moment to hope that it will be a success and gain popular enough acceptance to have device drivers and enough applications.

Maybe in the future, Open Source will focus on a truly modern Operating System without any of the baggage of the old UNIX era. Or maybe I ask for too much.

I hope this is not seen as trolling Linux. As a Linux/*nix user, I sometimes long for a far more simple Operating System that isn't either Windows or *nix/Linux and just worked without too much fuss.

Oh and Macs don't come into the picture. I'm talking about OSes for the PC.

Surely the *nix philosophy is all about flexibility, and chaining together tools which do one job. GUIs & sound systems are complex & I think the multiplicity of sound software is a reflection of this. A lot of the things you mention are either/or, like ALSA & OSS. There is a replacement for X coming, called Wayland, which acts directly on the hardware. I think a very small lightweight Linux system can be built that has a GUI & everything else - Damn Small Linux for example. Backwards compatibility is always an issue of course.
Don't forget the new smartphone/touchscreen etc. OSs as well - the Desktop PCs position as the dominant computing device & paradigm is under threat & its days are numbered.

I think what works with Linux for me is that I have choices. I am not just taking and excepting what everyone else is doing just because they are doing it or because it is easy and requires little thinking or decision making. Choice is never wrong or bad. When we have lack of choices then we are in real trouble.

andywebsdale, I also disagree that the Desktop pc's days are numbered. They are far from over. Just because we have automobiles doesn't mean that it killed bicycles. They both just settled into their proper places. Maybe not as many people ride bicycles like they use to but there still are many who do. And today with the cost of fuel, people are going back to riding bicycles because it is cheaper over all. Can't wait until they start pushing for more bandwidth caps and the cost of that sky-rockets with mobiles devices and this whole cloud thing. With Desktop pc's at least the persons data is on their computer and does not require bandwidth to access.

I just meant that its days as the dominant device are over - that's why Gnome, Windows, Ubuntu etc have all remodelled their UI's in a dramatic fashion. Windows 8 in particular will shock a few people I think. But of course there'll always be a place for a desktop style device - it's my preferred mode of computing by far.The question is though, how do teenagers & kids feel about it? Change is inevitable, & ever faster it seems....

@OP -Your frustration with Linux is basically what Linux is all about. I personally think it'd be a whole lot better if certain things were integrated more tightly (like the DE) but then what I see as better, another user may not. I think Haiku is a step in the right direction in terms of integration and simplicity of design but it's too restrictive for many people who like a bit of variety It's nice to see how far it's progressed; I remember when it was still called "OpenBeOS" or something like that

@Amdx2_x64 - Your analogy with bicycles vs PCs isn't quite accurate IMHO. I think it'd be more like horse & cart vs cars as a mode of daily transport. In which case the former is virtually non existent in industrialized countries. I personally don't think desktop PCs will die out for a while, but I do think they will be gone by the time I'm one foot in the grave.

<<<I don't mean to hijack this thread. But this is all more or less part of the same thing. Linux is having its day on mobile devices which I believe will eventually, in the future, benefit desktop PC's greatly.>>>

I agree. People are really going to be in for a surprise with Windows 8. After trying that I actually started to like Gnome 3, lol. I think too many people are counting on the cloud being the next greatest thing. I see it as a lot of marketing bull. Cable companies have shown us the end result. Cell phones have shown us the end result as well. They get people in, they kill real competition and the cost sky rockets. The cloud is a bad idea for many things. But all will see soon enough. Your data is not yours. Governments can get that data and companies will just hand it over. Look at Yahoo, Google, Microsoft, etc recently. Then if you're found to have anything that is illegal or questionable then off to jail you go (look at the copyright laws these days.) Corporations can get a profile on you so they can target you with their products and services. Governments will get a profile on you to watch and see who the potential trouble makers are or might be (even if you are not even close to being such.) And then along come those who hack into accounts and steal your information and data. Not to mention off line criminals who watch for patterns and when you are not home by what is posted online.

How can people not see this disaster they call the cloud coming. I am not creating anything new. These are things that have been in the news recently that governments, corporations, hackers, off line criminals have already started doing.

Desktop PC's will never die. They may go slightly into the background for a bit, but thats it.


Edit: I would compare horses and buggies to the abacus rather then the desktop computer.

basica, I've been using Linux as my primary desktop for nearly 10 years now. Seems amazing when I think of it myself.

So calling it "frustration" is a bit over the top as I've been more than happy with Linux all these years. I am just a bit annoyed that there is a lot of historical baggage and duplication of effort as a result of efforts to keep Linux/*nix up-to-date.

Simplicity is appealing. I am not against flexibility but against the historical baggage brought by a result of diverse efforts to achieve similar goals. Maybe it's inevitable anyway.

Back to your original post. When you say reinventing the wheel. After some thought I agree with that. I think many have been so caught up in not being MS Windows but being a better version of MS Windows that some have gotten into a rut. Of course having hardware and software companies mainly focus on MS Windows hasn't helped at all. With just hardware alone a lot of time and energy has been put into getting it to work with Linux, which could have been used in other areas.

I believe android is a candidate of breaking historical baggage.

It is a consequence of opensource being "chaotic evolution". To get KISS solution, you need one guy with "vision" of final product that would lead the whole thing.

Umm... no. Android is not linux. It uses fork of a linux kernel, and the rest of software is not GNU.

Desktop PCs are unlikely to disappear. You can't comfortably write programs on a smartphone, you know, plus handhelds have limited "power".

@OP
you already have a non-X non-windows OS just get to a command line and type "init 3"
there were a few vary few apps for svgalib and almost none at all for the frame buffer display

one great thing about having layers is you can fix the upper layers from the lower layers
I ran in to a problem with an old laptop that forced me to install X from source
had X been integrated in to the kernel I would have been shit out of luck and had to throw out the computer as unusable

there are times that I drop to run level 3 because it works better for text based things like working on the system it self or building something from source
I would not even think of building a kernel with X running
and because X is a layer it dose not even need to be installed on a file or web servers
that pretty display is a resource hog even when it has been swapped out by opening a virtual terminal


the sound system is more of a matter of evolution ALSA has replaced OSS rather than being an alternative the OSS driver that is still around now is for backwards compatibility
the DE and wm sound systems are more or less for the use of the DE or WM to use in there operation AFAIK the only apps that use a DE's sound system are things like the software suite for KDE

unix was developed at a time when computer time coast hundreds of dollars an hour so there was a deep , pressing and profound need to get the most out of an expensive resource at least 15 years of thinking and experimenting went in to coming up with the over all design of unix and I think they did a damned good job of it

in short there is no way in hell I could disagree more than I do already

rob.rice, I think you are going too much into specifics. This post is more a discussion on the general Linux/*nix historical development than technical aspects.

I don't dispute the technical aspects and I'm well aware of the advantages of Linux/*nix way of doing things.

See, the whole point of this discussion is about why this kind of questions exists anyway and I am not even questioning why Linux developed this way. I fully understand the historical baggage and the reasons for them. And besides only a person who is so interested in tweaking the OS would actually be interested in these details. And I am interested to an extent otherwise I wouldn't have used Linux for almost 10 years now.

A non-*nix OS of the kind I am talking about, would just work out of the box on a particular hardware or it wouldn't. Black and white. So you don't waste too much time into investigating issues or finding out why X.org or whatever would compile or not. And many Linux users including myself enjoy these kinds of explorations or we would simply not use Linux/*nix.

BUT... sometimes, viewing things as an end user, I yearn for simplicity of a single layered, stable, consistent OS that just works as what it is meant to do and doesn't offer so many bewildering and sometimes conflicting ways to achieve the same end.

I never thought I would agree with SigTerm , but I have to concede his view that a single, focussed vision is important.

Well, I guess the advertisers who sell android devices need to do their homework. Sometimes they use the words android and linux in the same sentence.

I knew android was a fork kernel but did not know about the GNU tools not being included. :/

Yes,very true. There are those who think it pedantic when people speak of GNU/Linux rather than just Linux, but Android's existence highlights the fact that, while the kernel is important(OK,a vital necessity) the rest of the OS needs to be fundamentally free also. GNU is a fantastic achievement & certainly without R Stallman's vision & massive technical input the distros of today couldn't exist in their present form.(Unfortunately Mr Stallman isn't as media-friendly a character as Mr Torvalds to put it mildly , so he doesn't get the plaudits he deserves)

Puppy Linux may fall somewhere in there.