The title of the article seems completely wrong to you? Naturally it would, when you daily read something like this. But I do state all this stuff is being a big mistake, if not worse. I am sure, that Linux is now close to extinction, and still is getting closer and closer to the point of no return. Before we enter the discussion, please accept the following: in this article I’m not giving any opinion on topic of software freedom or openness. Neither I discuss the pros and cons of UNIX way and WYSIWYG. All the words concerning these issues are just describing the situation, but never expressing any attitude.
Now the Linux-based operating systems are rising. Linux is being run on numerous systems from Internet servers to employees’ and home users’ desktops. More and more companies (or even administrative bodies) are moving to Linux. Many well-known software vendors abandon their previous system software projects in favour of Linux (as Novell and PalmSource did) or enter the OS market with new Linux-based solutions (just like Oracle). Many of software developers or even software houses (like Sun) port their core applications and services on Linux. Linux is widely recognized as successful project and reliable business platform.
The Linux advocates are trying to make us think this process shows the success of Linux and respectively UNIXes victory over Windows and fellows. Is it really so?
My answer is: no.
Before we enter the discussion, please accept the following: in this article I’m not giving any opinion on topic of software freedom or openness. Neither I discuss the pros and cons of UNIX way and WYSIWYG. All the words concerning these issues are just describing the situation, but never expressing any attitude.
Linux was just one of the numerous projects, which happened to rise due to being distributed as Free Software and supported by FSF. This sort of publicity made it possible to start several commercial projects (namely Red Hat, SuSE, Mandrake and some Debian derivatives) based on Linux. The rise of free-of-charge Linux-bases system attracted the views of software vendors who started sharing the benefits of Linux’s publicity by contributing to the project.
Being opposed to Microsoft’s monopoly over the system software market Linux had to keep competing with its rival. While the customers believed the desktop software to be exactly what Windows was, the vendors started to invest the projects that shared theses views. Ever since Linux was becoming more and more like Windows, providing the same user experience and utilities resembling those of Windows and fellows.
The GNOME project is a good example. Being initially intended to provide the IBM OS/2 user experience it gained vendor attention. When the project’s officials started to state that the next goal of the project was the Windows UI, it became the default desktop in some commercial distribution (including Red Hat), and the GNOME adoption fastened when the developers starter to decrease the features amount, making it really no more difficult to customize than Windows’ Shell32 UI.
The situation is fundamentally bad. Do you want to know why? Stay tuned.
The matter of OS choice
The user would never mind the operating system he uses. The computer is simply a tool for completing the user’s tasks, so the only valid factor to be taken in consideration while choosing the operating system is the default set of approaches. That’s to say your system must provide you with the instruments most compliant with your mindset and your way of acting.
This time we have another issue that is taken in consideration by some of us: the freedom, that is delivered to us by the software. Those who believe that the software they use should be easily customizable and modified choose among the open source operating systems. Those, who think that the user must be given the freedoms to use program any way he feels appropriate, to share it without charging the fee for such sharing and to make his modifications of the program accessible to general public, has to choose among the so-called Free Software operating systems.
Currently only several UNIXes are both Free Software or open source ideology compliant and stable enough to be used in mission-critical systems, so no choice is left for the open source and Free Software followers. But this situation is starting to get better: the ReactOS, Haiku and GNUstep projects are being actively developed, so we are about to see the Free Software operating systems in style of Windows, BeOS and MacOS X respectively.
This time we have two different styles of operating available: WYSIWYG and non-WYSIWYG (I can’t recall any good all-known word for it and don’t want to introduce my own term). The first one is the natural domain of Microsoft Windows, Apple MacOS (including OSX) and BeOS (now ZETA). The later once used to be the default UNIX’s style. Things changed dramatically since.
The WYSIWYG and traditional UNIX way to accomplish the task are incompatible. WYSIWYG (acronym for What You See Is What You Get) introduces user to a graphical environment, where the result of every manipulation is displayed as it would be seen in result. For example you are shown the document and you see how it would look like. But there is no room for logical structure of document in the case of WYSIWYG software. The other way puts it different: you are presented to a plain text document with the markup describing both logical structure and formatting, but you are not given any idea about how that all would look on paper.
I don’t know for sure whether it’s really more handy for an average person, or it just has much better publicity, but WYSIWYG is currently widely recognized as the preferable way, making WYSIWYG implying operating systems most popular. As I mentioned above, commercial Linux distributions’ default UIs followed the successful WYSIWYG styles. Being promoted as cheaper alternative solutions they pretend to give user the same level of usability and productiveness as Windows while being similar enough to make user transitions nearly seamless.
So, we are having Linux in position of WYSIWYG-implying operating system. Well, MacOS X in its turn is actually a desktop environment based on UNIX-like Darwin operating system. But the way OSX wraps the underlying OS makes me feel comfortable while excluding OSX from UNIXes list. It would even be reasonable to state that if Apple decides to move OSX from Darwin to some totally different non-UNIX-like platform, the users won’t feel the difference. But Linux is really UNIX, and unlike MacOS X Linux doesn’t really try to hide away the UNIX nature of the system, although the percentage of UNIX-styled software decreases dramatically.
Anyway, the idea of Linux as UNIX-based Windows clone feels optimistic for everyone except for UNIX way zealots. But if we take a closer look, we’ll find that this model leads Linux to ruin.
The unfair race
Why would Linux loose trying to provide the same user experience as Windows, but at lower cost?
First of all, you can’t copy the thing you’ve never seen. Microsoft’s plans are changing all the time, some features are found only at release time, so they would hardly find their way into Linux UI in short term. And if they are technically not so easy to implement, they would be always outdated, just like Mono, which being sponsored by Novell is still a .NET 1.x implementation, having incomplete 2.0 support in development branch, while some of current software depend on .NET 3.0!
The reader could object me that any new feature needs to be adopted before it is used, so this time of adoption will be used by developers to release the new feature right when users would start using it. So true, but the way the destiny of the feature is largely dependent on the developer of the very first version.
The second reason for Linux to loose is that the work needed to provide this kind on Windows-on-top-of-UNIX is close to literally developing two operating systems with different userland. And the further we go, the more trickier the system gets. SysV- and BSD-styled init scripts are being replaced with event-based system in Ubuntu, different other over complicated projects are developed to make the UNIX-like Linux base system resemble Windows.
The amount of work rises, the level of overcomplication rises, the number of layers between user and hardware rises… You don’t need to be a programmer to understand, the more code you have, the less easy it is to maintain it, to update it, to modify it. While no complexity barrier being seen this time, we are still to reach it some day.
The third reason is the user base of Linux. Until some very recent time the most part of Linux userspace developers were working to make some software they needed for their own systems. And they didn’t think about the ease of use for Windows users, about Windows-like interfaces and so on. They were just making the tools for their means in their operating system, which was still UNIX those days.
Now they are getting out of touch. They stop updating their Linux boxes or even migrate to less Windows-ish flavours of BSD. Many of really smart projects that are really needed in UNIX system Linux might be stopped updating. The short term tactical victory of Linux still makes the loss of UNIXes evident.
Nothing could be stopped now. Microsoft promised not to sue Novel’s clients, and surely it won’t, because the more Novell relays on this promise, the further it goes to the Microsoft territory, where no one but Microsoft has the chance. On seeing all of this Novell entered the deal, preferring to gain the short term outcome. The vendors have all decided for us.
So, what is coming next?
First of all, the Linux distributions will split in two groups: those relaying on UI and Windows-ish tools and others relaying on UNIX way utilities and command-line interface (which is actually the most powerful interface in most cases). The first group being supported by commercial vendors will continue improving WYSIWYG abilities of Linux, while the second group will instinct, the users will switch to flavours of BSD, Hurd, Plan9 or whatever.
After some time the Linux development will turn too expensive for Linux vendors be competing with Windows. Some of vendors (Oracle for example) will be forced to leave the system software market, some may be die out, the rest will switch to some alternatives (ReactOS has very strong chances to replace Linux).
By this time the small amount of pure UNIX open source systems will occupy some very slow share of OS market, being supported by classic UNIX hackers, geeky communities and FSF. They will take the place of Linux of the early 200x.
This was the optimistic vision, and here we go with pessimistic one:
Windows Vista will appear to be the failure the Microsoft won’t overcome. The Linux quickly becomes the main platform for all sort of commercial software.
After a period of time there would appear different closed source superior substitutes for some of subsystems. Being not too innovative but still incompatible, they would be adopted by different vendors in random order, making the market as divided as it was at the Microsoft time.
Without any valid competition the market will keep stagnant until some new Microsoft-like player won’t defeat the weak Linux vendors one by one and occupy the Microsoft’s place.
Any way, no good news to come from Linux any more.
About the Author:
Dmitrij D. Czarkoff is a Russian intellectual property and insurance lawyer, spending his free time on free software advocacy and UNIXes promotion. OpenBSD user, if you would like to know.
If you would like to see your thoughts or experiences with technology published, please consider writing an article for OSNews.