This is a commentary. From a Linux user who does tech support for Windows users and works in the real world of a corporate Windows network environment. 2004 has been touted by many as the year of the Linux desktop. Indeed with the backing of IBM, Sun and now Novell, the business world looks like getting a serious Linux desktop contender. But has Linux on the desktop really got what it takes?Editorial Notice: All opinions are those of the author and not necessarily those of osnews.com
Linux as a hobby OS is great. You can make a personal statement about your independence and hacking ability, and sleep better at night knowing you have a superior OS without so much as a nod to the great satan. Linux as a server is greater still. The bang for your buck, the ease of installation and reliability make it an ideal candidate for mission critical situations. But Linux as replacement for Windows on people’s desktops?
The Windows desktop is a highly developed and polished product. A Macintosh desktop is likewise, unfortunately OSX continues to only be available on fairly expensive hardware made by only one company, Apple.
Despite all of the deficiencies of Microsoft and Windows, the fact of life is that refinement over the years has produced a current Windows desktop which is slick and functional. In addition, almost every piece of consumer software will run on Windows and almost every item of hardware will work with Windows.
So if Linux is to compete in some way on the desktops, clearly a number of the key features of a Windows desktop need to be matched. Mainstream computer users have years of experience with Windows, forcing experienced users to sacrifice some of the mod cons of Windows isn’t going work. This is one of the reasons why Linux in developing economies makes so much sense.
What are these “mod cons” that make a Windows desktop?
Aside from the GUI and point n click, there are some obvious hallmarks. These are:
* Uniformity of appearance
* Consistent behavior
* Ease of use
* Ease of installation
* OS transparency
There are bound to be more, but these strike me as the most obvious characteristics. The question of abundance with regard to software is another issue. As many Linux advocates have observed, a few killer game applications would do more for the cause of Linux than possibly years of more quiet development. I suggest that although Linux software development proceeds along different lines, ultimately commercial products will arrive when the market appears.
So what do I mean by these desktop characteristics? In the main, they are features that Windows users, and by extension, the rest of us, expect when using a modern PC.
Uniformity of appearance – this is either a no issue or a big one, depending on your point of view. Some people will suggest that in fact Windows doesn’t do very well in this department compared with say OSX, however whilst the more modern Windows systems may not have the same eye-candy as OSX, they still have a uniform look and feel. Unfortunately, an historical precedent has given rise to a diverse and at times competing window environment for Linux, the end result is that today’s Linux desktop can look and feel like a dog’s breakfast.
Consistent behavior – again the quick wits will say that the consistent Windows behavior is all bad, which might have been true some years ago but is far less true now. To be precise, this characteristic is closely related to appearance. For example when an icon of some sort appears on the desktop, then clicking on it with the mouse will produce a predictable response from the system. Users learn to associate and differentiate icons and their functions and become comfortable in the security that these things tend not to change. Again this has not always been a strong point with Linux desktops and to a certain extent highlights the different approach that Linux developers have taken.
Ease of use – these two feature combine to provide a degree of ease-of-use, but there are others. Desktop GUI’s reflect a certain thought process, if that process is one that shares a large area of common ground with users either by way of computing experience (ie Windows) or personal politics (say OSX) then users will be able to instinctively guess an action to achieve their desired outcome. By its very nature Linux will be forced to emulate Windows in this regard as the established desktop Linux user base is diminutive. Unfortunately emulation will always leave the user comparing the copy with the original. Ease-of-use extends to other areas; for example Windows users can easily use software and hardware because it is so widely supported and available.
Ease of installation – possibly one of the weakest user characteristic for Windows but still one that encompasses significant strengths. For users, installing a program is simply a matter of double-clicking the installer. Window’s uncompromising approach to a PC’s hard disk utilisation is an example of how a lack of choice and real options creates an easier job for end users. Most users don’t really care how a hard disk is partitioned or how it boots provided it works without users having to get involved. Live Linux distros like Knoppix have a lot to offer here. On the one hand they offer a zero pain demonstration of Linux to Windows users and on the other hand they also offer a relatively easy installation process. Most modern Linuxes also offer relatively easy installers, however the addition of software to a Linux system is often more problematic when compared to a Windows environment.
Reliability – which was for a long time a very sore point for Windows, has become one of its strengths. Yes Windows crashes, but then so do Linux boxes. Linux servers have probably got the edge, but on the desktop its not so clear cut. Complicating the desktop scenario is the user. Users tend to do funny things to their computers and are often guilty of self harm. Whilst the security of a Linux system may be more stringent and thus prevent major problems, examples such as Lindows and root passwords highlight the potential here for a repitition of the Windows scenario.
Aside from such concerns there is another component to reliability. Consider this example, Jane buys a new digital camera for her home PC. She uses Windows at work and at home she has a rent-a-box which has XP and all the usual stuff. She may have problems with viruses and she might have security concerns about the internet, but mostly she wants to be able to do her homework and now she wants to hook up her new camera and print some pictures. This is not and should not be a drama, most of the time for people with Windows and some of the time for Apple users, this type of scenario IS NOT difficult. Jane checks that her new camera is compatible with Windows XP and when she opens the box there is a shiny new cd to help her out. In this sense, Jane can rely on Windows to do her job. Simple scenarios for everyday users that can be reliably done. Linux users can do some things as easily and as reliably as their Windows friends but clearly not as many.
OS transparency – for many Linux advocates, this is an anathema. They delight in demonstrating that the command line is lurking, easily accessed with a few keystrokes. Granted, it is a tool that professional users derive great benefit from, not so the average desktop user who pays homage to the twin altars of point-n-click and GUI, and with good cause. Desktop users want to check their mail, write a report, browse the web or run a million other applications on their computer. The OS is a sideshow to the real action, doing stuff on your computer means applications and furthermore, running those applications doesn’t mean getting involved with the inner workings of your operating system. If an application is available for both Windows and Linux then users have every right to expect that using that application on either platform does not involve a complete re-education. In other words, Linux will be successful in this regard when users are more or less oblivious to the fact that the OS on their computers is in fact Linux and not something else.
Linux has great potential as a desktop computer with some definite advantages over Windows but until Linux matches some of the key features of the Windows desktop then any mass adoptions are likely to limited to specifically targeted niche markets and newly evolving ones.