To view parent comment, click here.
To read all comments associated with this story, please click here.
I don't believe it has anything to do with choice.
Linux is just a kernel. The OS is Mandriva, Ubuntu, Linpus, Xandros, Fedora, XP, Vista and OpenSuse.
Mandriva packages are not compatible with Ubuntu packages, yes. But neither are Windows XP packages with Fedora ones. Here it's not a problem of unity. Mandriva is united, and so is Fedora, Ubuntu and Windows XP.
And yes, 1000 copies of Windows XP are sold when one Ubuntu is installed. You believe it's because Ubuntu is not compatible with Mandriva? I don't think so. Look at reactos. reactos is (at least aims at) compatible with windows. Do you know reactos? How many people know Reactos?
The problem is elsewhere in my opinion.
I'd point out the egg and the chicken problem. Applications are not ported to Ubuntu or Fedora because almost nobody use them, whereas everybody use Windows. Everybody use Windows because the apps are available. People know windows and they buy Windows because their neightbour use Windows.
Ubuntu and Mandriva have a compatibility layer (wine), which is almost stable at version 1 (release candidate 3 as I speack). Reactos is fully compatible. Most Windows apps work on wine and therefore on Ubuntu and Mandriva. And yet, people don't know it.
The biggest problem here is marketing. Mandriva is free. Mandriva has not the same money as Microsoft to advertise and will never pay Acer to put their OS on their computer. Another chicken and egg problem. The rich get richer and the poor get poorer. This always has been and this always will be. The monopoly sustains itself and nothing can change that.
Edited 2008-06-05 08:58 UTC
I can't believe that you're using ReactOS as an example of the failure of open-source to get a userbase. ReactOS is not "fully compatible" as you state, and it's currently alpha-quality software at version 0.3.4. It tells you right there, in bold, on their web-site that it's not for everyday use. Perhaps you're claiming the same for Wine, but regardless of whether it's at version 1.0 or 3.0, it does not run every Windows application flawlessly.
There are many reasons why it's taken Linux a long time to gain a userbase, but your argument that "it hasn't happened yet, so it's never going to happen" is just a big blanket denial of all future developments. If that were fair, we wouldn't even see the mainstream Linux devices that are appearing now.
Firstly, don't nitpick me about Linux is only the kernel, I'm well and truly aware of it and was probably using Linux before you were (circa 97). I used Linux as a generic term, because that's what most people do (mistakenly I do admit).
Yes, I know of reactos - alpha quality software, probably always will remain alphaware due to the small numbers both using, and developing it.
Microsoft is part of the reason why Linux is not seeing further adoption, but only part of the problem.
Get out and talk to ordinary users - ordinary people. You will see that the vast majority of them both dislike computers, and do not understand both the technology, and how to use them, at least to varying degrees. Herein lies the problem. When someone doesn't understand something, they dislike complexity, or choice. Choice means more things to think or worry about, or to try and understand. When people just want to use something to get the job done, they don't want to have to think about it. The trained monkey scenario is sadly quite true in the average home/workplace.
All this choice - Ubuntu, Fedora, Suse, Mandriva and so on and forth only splinters the Linux machine. The same applies to package management systems etc and desktop environments. Linux needs to adopt the freebsd approach - one system. It'll make things easier for adoption by system manufacturers. At the moment, Linux is a nightmare to ensure compatibility. Imagine you're a super application writer like Adobe. Getting a Linux release of Photoshop will probably not happen, because making it install on a current Linux distribution might be easy, but on all of them, with different system locations? Doubtful, complex and expensive. Making it work on older Linux systems with different versions of glibc? Forget it. Whilst Windows isn't perfect, it's a helluva lot better than Linux in this respect. Something made for Vista will almost certainly (in most cases) install on Windows 98 and vice versa.
Dave
Is the goal to kill Microsoft?
Is the goal to take the market?
Is the goal to have great software that can be studied?
For many, the goal is simply good software without synthetic limitations.
"Linux" is the commodity part which many have chosen to build there own specialized distribution with. Turns out, it supports a boggling wide array of different uses.







Member since:
2006-12-16
And that is both Linux' strongest and weakest points. Choice is good, but only if your market is capable of making use of the choice. Your average computer user doesn't give a flying **** about choice, they just want it to work. Too much choice is confusing, and therefore undesirable.
All the people pushing Linux just must not be able to look at computers and software from the eyes of the average user, cos Linux keeps making the same mistakes.
One DE. One package management system. One distro. That is the ONLY way Linux will go forward in the market. We'll still be here in 20 years with Linux having 3% of the market I honestly believe, unless things change.
Dave