AH Challenge: Linux has a major market share by 2010
- Thread starter Aranfan
- Start date
Overall, it is second only to windows variants if you count embedded devices (nearly every router uses Linux, and Android now is the second largest smartphone OS) and servers (something like 70% of the Web runs Linux). So, it's only really on the desktop where it lags, as third after OS X.
The PoD is thus simplistically easy, have Apple adopt the Linux kernel for OS X instead of NeXT's BSD variant. Done.
The PoD is thus simplistically easy, have Apple adopt the Linux kernel for OS X instead of NeXT's BSD variant. Done.
I'm not so certain... Apple adopted NeXT's BSD variant because they adopted NeXT outright, and of course NeXTStep predates Linux... plus, the GNU General Public License would put much more of Apple's code out there, which doesn't seem to be something Steve Jobs would easily accept...
It seems to me that some potential (but probably subverting the question) answers are:
1) Kill Apple in the 90's. A lot of the Apple people would probably end up going MS for lack of options (which could disqualify this under the "widely acknowledged market share") and Linux is kind of different from the Apple mindset, but a decent percentage could end up looking for non-MS alternatives, and this is when GNU/Linux becomes more prevalent...
2) Have Microsoft focused on Xenix to the exclusion of Windows NT or even OS/2... Microsoft would then suffer many of the disadvantages the other Unix vendors did with regards to Linux, and would not have as easy of a job combining its professional and consumer (DOS-based) lines without NT's baked-in backwards compatibility.
Problem is --Linux is not put out as a Operating System - aka DOS, Unix, NT, BEOS [original OS, not today's ???] or BSD.
Instead Linux is hundreds of Distros, composed of application packages wrapped around KDE or Grome Kernels.
Each Distro is made by Geeks, with their choice of Applications, with no real feel for Harry Homeowners problems adding, removing, or changing the applications.
What Linux need is to separate into the KDE-OS, and the Grome-OS, and have all the Linux people work on the Applications, Which are not finished till the Install program is inside the Application.
Instead Linux is hundreds of Distros, composed of application packages wrapped around KDE or Grome Kernels.
Each Distro is made by Geeks, with their choice of Applications, with no real feel for Harry Homeowners problems adding, removing, or changing the applications.
What Linux need is to separate into the KDE-OS, and the Grome-OS, and have all the Linux people work on the Applications, Which are not finished till the Install program is inside the Application.
When BSD hit it's legal problems with AT&T, Lunis Trovald [early computer Geek] and friends set out to write a Kernel that would not have the legal problems. This was KDE.
However there were personal problems with some of the others, so the Group Split. The second Group went on to Write the Gnome Kernel.
IIRC Linux comes from Lunis' Unix. Lunis has said that if it hadn't been for BSD's legal problems there never would have been Linux.
However there were personal problems with some of the others, so the Group Split. The second Group went on to Write the Gnome Kernel.
IIRC Linux comes from Lunis' Unix. Lunis has said that if it hadn't been for BSD's legal problems there never would have been Linux.
I've experimented with Linux for quite some time, and feel reasonably comfortable with it most of the time, but I can see why it hasn't been widely adopted.
The power and weakness of Linux on the desktop is the diversity. What we normally call Linux is really a mix of (1) The kernal, which is actually fairly standardized, though not always (2) The GNU utilities, which are extremely powerful but mostly geek oriented and command-line based, (3) The graphics interfaces, of which there are quite a few. The most common ones are KDE and Gnome, but there are several others. When you go from one Linux computer to another you can end up with very different interfaces, and a steep learning curve. Gnome and KDE may not be quite as different as Windows and Mac, but they come close.
Add in the fact that the various distros customize which utilities are included even within a graphics interface, and you have a lot of confusion, which adds up to people getting turned off to the OS.
As to how you get to a major market share for Linux: I don't know. A major computer manufacturer getting into a major dispute with Microsoft before it was too entrenched and dumping Windows for Linux might do it. Gateway may have toyed with something like that after they acquired Amiga. They were supposedly trying to build an OS with the Amiga interface on top of an embedded Unix-like OS. I forget which one it was, but I don't think it was Linux-based.
Actually, I think a variation of BSD might be a more likely desktop operating system competitor. As I understand it, the BSD license makes it easier to build proprietary systems on top of the OS. I could see the big US manufacturers like HP, Compaq (when it was still independent) and Dell deciding that sticking with Microsoft meant that the bulk of the profits went to Microsoft and they would be reduced to makers of generic boxes. In that case, they take BSD, put a Windows-like interface on it, and try to compete based on that. Ultimately I think the best they could do would be slightly better than Mac-type shares, and they might not do that well, but I could be wrong. If all the big US manufacturers went with non-Windows systems at the same time early enough they might be able to pull it off.
Linux does have an opportunity with tablets if tablets continue to take off. Microsoft so far doesn't seem to have a clue how to make a successful tablet. A unified Linux that stays unified might do well in that market.
The power and weakness of Linux on the desktop is the diversity. What we normally call Linux is really a mix of (1) The kernal, which is actually fairly standardized, though not always (2) The GNU utilities, which are extremely powerful but mostly geek oriented and command-line based, (3) The graphics interfaces, of which there are quite a few. The most common ones are KDE and Gnome, but there are several others. When you go from one Linux computer to another you can end up with very different interfaces, and a steep learning curve. Gnome and KDE may not be quite as different as Windows and Mac, but they come close.
Add in the fact that the various distros customize which utilities are included even within a graphics interface, and you have a lot of confusion, which adds up to people getting turned off to the OS.
As to how you get to a major market share for Linux: I don't know. A major computer manufacturer getting into a major dispute with Microsoft before it was too entrenched and dumping Windows for Linux might do it. Gateway may have toyed with something like that after they acquired Amiga. They were supposedly trying to build an OS with the Amiga interface on top of an embedded Unix-like OS. I forget which one it was, but I don't think it was Linux-based.
Actually, I think a variation of BSD might be a more likely desktop operating system competitor. As I understand it, the BSD license makes it easier to build proprietary systems on top of the OS. I could see the big US manufacturers like HP, Compaq (when it was still independent) and Dell deciding that sticking with Microsoft meant that the bulk of the profits went to Microsoft and they would be reduced to makers of generic boxes. In that case, they take BSD, put a Windows-like interface on it, and try to compete based on that. Ultimately I think the best they could do would be slightly better than Mac-type shares, and they might not do that well, but I could be wrong. If all the big US manufacturers went with non-Windows systems at the same time early enough they might be able to pull it off.
Linux does have an opportunity with tablets if tablets continue to take off. Microsoft so far doesn't seem to have a clue how to make a successful tablet. A unified Linux that stays unified might do well in that market.
My understanding is quite different: Linus made Linux as a research project and an alternative to Minix, which was another research project whose creator, Andrew Tanenbaum (I hope I have that name right), refused to add a lot of updates the community wanted because he wanted to keep it understandable for his students, as he taught a class on operating systems. Linus adopted a lot of GNU utilities, and the project was found by the GNU project, whose own kernel (HURD) was having difficulties in being programmed, and could not adopt BSD because of the AT&T/USL lawsuit. But early Linux was command-line: KDE and GNOME were two projects, to bring a GUI... KDE was the original, but it was based on what was then non-free software, so GNOME was a free alternative sponsored by the GNU people. My understanding is that there is still one Linux kernel, just different software suites that run on it.
Flocculencio
Donor
This. This is it- so long as Linux doesn't come in a form that the majority of casual home users can just start up and use, it won't be a successful mass market product.
Exactly true. From an user standpoint, however, darwin (the OSX kernel) and linux use the same command shell (bash). The two shells are functionally equivalent, even if the kernel code is radically different in places. Many programs can be compiled to run either on linux or on darwin. However, code compilation is not something that a casual user would want to do. In reality, darwin/OSX most often uses software designed for the darwin kernel and the OSX desktop environment.
The most popular desktop linux distro today is Ubuntu, which has about 50% of the market. "Standard" Ubuntu uses Gnome, but one can have Ubuntu with KDE (kubuntu), XFCE (xubuntu), L
E (lubuntu), and even Fluxbox (fluxbuntu, no longer in development). Most Linux users never go outside of Gnome or KDE and most of the other desktop environments use Gnome or KDE libraries. In effect, there are only two mainstream desktop environments for Linux.Soon Canonical (the distributors of Ubuntu) will introduce a new desktop environment (Unity) and window manager (Wayland) to replace the aging xwindows. While mostly open-source, this DE and windowing engine will take Ubuntu much farther away from the other xwindows-reliant distros.
I could see desktop Linux taking off if:
1) Canonical takes Ubuntu/wayland off into a semi-closed-source kernel and markets the new kernel+window system combination aggressively to netbook and low-end PC manufacturers. This might be a good solution for cheap desktop PC's that use netbook processors like the Intel Atom. Perhaps Canonical could license some code from Apple and design some application cross-compatibility into their semi-closed kernel. The Ubuntu kernel would be downwardly compatible with most Linux programs as well. The semi-closed or closed Ubuntu kernel and DE could appeal to those who want OSX look and functionality on cheap PC hardware (and aren't willing or able to make a Hackintosh).
2) Ubuntu could market its closed source linux-like kernel/window system and continue to develop its Linux distros on a less formal basis. It could continue to make money off of Linux enterprise support as well as take in revenue from its new closed source kernel.
Even if Canonical drops development of its "traditional" linux/xwindows based open-source distributions, the Debian community would probably pick up development again. After all, many of the Ubuntu utilities (particularly APT) are Debian-sourced developments.
There will always be Linux for nerds. Question is, can Canonical create a semi-closed system which provides downward compatibility with the Linux kernel and tap into the OSX software market?
Last edited:
Well, they can now. Ubuntu is only one of many distros that are fairly easy to use, just as easy as Windows or Max OS X, and has been up there for several years already. There is a bit of a learning curve if you're used to Windows or Mac OS X and haven't used anything else. However, if this is the first computer interface you've ever used, the learning curve is going to be just as much for Windows, Mac and Linux.
The problem is that the computer manufacturers are getting benefits out of the commercial marketing schemes of Windows and Mac. If they aren't outright paid to sell machines with those OSes on, they get most of their business from people who see Windows and Mac commercials and want one of those or the other.
A POD would require someone, somewhere back there, choosing to spend money on marketing. Some major corporation which decides to sell inexpensive, but high-quality computers by releasing its own, really user-friendly distro of Linux. At the same time, it's funding the development of this distro, which only improves Linux in general.
An alternative POD, however, would change the PC market completely. Back in the very early '80s, before DOS and Apple really began to take off, you could easily contemplate a world where the OS wasn't important, it was the manufacturer.
The idea of buying a machine because of a specific interface is not exactly ubiquitous in commercial devices. You don't buy your washing machine or oven based on the color of the dials.
Most people don't buy their cars because of the location of the gear shift, the number and appearance of instrument readings, or where the seat controls are. You might prefer a car based on some of these factors, but you wouldn't turn down a car with better performance/cargospace/features just because their dashboard wasn't specified and designed by Car Interface, Inc.
If a car analogy isn't good enough, look at cell phones. Each device, sometimes even with the same brand, has a different interface. Whenever you get a new one, you need to learn how to do all the things you want to do with your phone all over again. However, people choose phones based on features, or whatever discounts the service providers are currently operating.
So, if early PC manufacturers built more in-house operating systems, or contracted out their operating systems to companies that were willing to reprogram them according to the manufacturers required specs, then operating system wouldn't be as important as manufacturer.
Of course, this isn't the way computer manufacturing and operating system development works.
Last edited:
pnyckqx
Banned
Probably the best shot at that would have been a distribution originally called Lindows, and now known as Linspire, until it was bought by Xandros. Unfortunately, legal problems and management decisions prevented the OS from being all that it could be.
The difficulty of Linux was never learning to use the GUI. KDE is similar enough to Windows, just easier to customise.
The difficulty is that proprietary hardware developers refuse to write Linux drivers for their products. At this moment, it's hard to get a Broadcom wireless card working in Linux, depending on the chip set.
A good POD would be for HP and others to begin their support of Linux early. Probably about the time that Microsoft was forced to go to Linux servers when it's update site was hacked a few years ago.
If drivers are readily available, there is far less difficulty migrating to Linux, and far few malware problems in the world at the moment.
While security by obscurity is legitimate, it is damned difficult to write a virus that would work against a Linux OS.
This forget (maybe, I have quickly read all) one major computers market player of 'recent' (died out by mid 90s...) past... Amiga.
Do you mean huge market share of home desktops? Frankly, Linux was already dominant in servers by 2010 OTL.
Distros like Ubuntu are making inroads into business because they are semi free and are not that hard for Windows trained people to use.
Good POD to get Linux to have a major presence in the home PC market would be to have Walmart switch to a customized Linux distro designed by Canonical. Maybe Walmart would make its own programs in super user friendly Wal-Ubuntu and sell these machines to the public for low prices.
Distros like Ubuntu are making inroads into business because they are semi free and are not that hard for Windows trained people to use.
Good POD to get Linux to have a major presence in the home PC market would be to have Walmart switch to a customized Linux distro designed by Canonical. Maybe Walmart would make its own programs in super user friendly Wal-Ubuntu and sell these machines to the public for low prices.