At first glance Ubuntu appears to be the answer to the prayers of Linux evangelists worldwide. It has a great website, great marketing, an enigmatic philanthropist leader, a devoted community and a philosophy which seems to mirror that of the wider free software community in stark contrast to its enterprise counterparts. With such a stellar resume one has to ask the question, is Ubuntu too good to be true?
The only real problem with Ubuntu is Ubuntu itself. A tested and bug fixed version of Debian unstable with a pretty installer, a splash screen and the Gnome desktop is hardly the 'revolution' which it is purported to be. For Ubuntu to upset major players in the desktop arena such as Microsoft and Apple they need to start behaving like a professional company and provide for the needs of their customers as opposed to what the company thinks they need.
Let’s begin with Gnome. If Gnome was going to be a Desktop that would create massive changes in the IT world it would have done so by now. In no way am I belittling the success and advances Gnome has made within Linux, rather it’s the simple fact that Gnome is old and no longer up to scratch. Users want a pretty, easy to use environment with consistent menus and one administrative 'control panel' type utility. This should provide a simple solution for all administrative tasks such as adding and removing programs, hardware and networks. Gnome is not the answer for the Windows savvy world and neither is the resource-hungry KDE.
Innovation and diversity is what makes Linux different and supposedly superior to proprietary OS's. However, with every major distribution opting to use one of two substandard window managers and every release becoming increasingly unimaginative and generic is there really any need for another bloated, windows-like version of Linux? I believe the simple answer is, No.
I could go on for hours about my personal dislikes for this particular flavour of Linux but as it is relatively indistinguishable from so many others this would be pointless. Instead I will outline exactly what I believe the open source community needs to develop in an operating system which will interest the rest of the world enough to ditch what they know and try something new. Many of these features are available in existing offerings; I will try to acknowledge this where possible. I want to emphasize that this is not intended as a feature-list for the 'best' Linux distribution ever, this is a list of necessary characteristics needed to truly reflect the Ubuntu motto; “Linux for human beings.”
The Desktop Environment is the first issue I want to discuss as it is an obvious requirement for OS success. Unfortunately in recent years it has proved wholly unimaginative and under-developed within Ubuntu releases. Although there are distinctions to be made between window managers and desktop environments they are outside the scope of this article and it should suffice to say that a distribution “'for human beings” would comprise a Desktop Environment as opposed to the latter.
The current desktop environments namely Gnome and KDE have done great things to bring usability and eye candy to Linux on the desktop. To their credit they were once considered revolutionary and well developed, however like any piece of old software architecture they have outlived there usefulness and their popularity is due to the lack of a comparable alternative.
For Linux to retain its reputation as the fast, stable, secure and freely available operating system it needs to continue to live up to all these qualities. Gnome and KDE as the current default desktop environment means sacrificing at least one of these qualities. Sadly this is making Linux a less attractive solution which is bad for the community.
With no current satisfactory solutions available Ubuntu should have either created their own Desktop Environment or, better still, invested some of its millions of development dollars into one of the projects such as Enlightenment, Mezzo or the Sun sponsored Project Looking Glass. These are Linux projects truly aimed at making big steps in innovation as opposed to stirring stagnant water. All that would be required would be to get the code for one of these amazing projects, add some polish, remove some options and make sure it has the following elements designed to help, not confuse the intrepid newbie:
A control panel which contains everything (!!) needed to administer the system via simple gui-based utilities with a consistent look and feel.
A consistent look and feel across all applications available via the package manager. If it doesn't look right either change it or don't include it.
Speed. We use Linux because we don't want to be forced in to constant hardware upgrades by OS manufacturers that have strategic partnerships with hardware vendors or, in the case of Macintosh, are the hardware vendor.
Aesthetics. If you want to beat Mac and even Windows it had better look better than Gnome. I can draw a parallel with cars; no matter how much innovation is achieved in an automobile design; no one wants to be seen dead driving a Fiat Multipla. To the pragmatist aesthetics and speed may seem mutually exclusive but you only need to try Enlightenment to know this isn't the case.
Stability. Naturally this is a very important consideration but oddly enough seems to be getting less and less attention these days. It’s great that an application crash doesn't freeze your entire system in Linux but that doesn't mean application crashes are acceptable on a regular basis. Unfortunately this appears to be the case even with many of the major distributions currently.
Secondly and possibly just as important is the package manager. One of the major advantages of using Linux is the availability of high quality free software. The advantage this provides for distributions such as Ubuntu (who use Debian as their parent OS) is that they already have a brilliant command line package management system available in the form of apt-get or aptitude. The gui front-ends for apt however leave a lot to be desired, a 'human being' using the Ubuntu package manager would likely find it less than useful when they type in a query such as 'Internet' in the search box and are presented with hundreds of options for everything from the Bind DNS server to the more obscure libiiimcf2 package which according to its description is an “Internet/Intranet Input Method Framework Library”. Very useful for its intended purpose no doubt but even more useful for convincing a new user to never ever fire up that package manager again…. ever. Luckily for this particular problem a viable solution already exists, the interface for Freespires CNR software warehouse is simply brilliant. Copy it and don’t feel bad about it. Freespire make it glaringly obvious that they are using Linux and related open source software to copy Windows in every way possible, in fact they used to be called Lindows!
The last characteristic required of a “Linux for human beings” is unfortunately the most difficult to achieve and by far the most contentious but it is also the most important hurdle for Desktop Linux in my opinion.
Software support. Human beings don't like change and cannot be expected to change every piece of software they use all at once. Therefore an operating system for human beings wouldn't require them to do so, unfortunately it does. It’s true that there is an abundance of excellent software for Linux that makes it possible to achieve almost anything within the realms of the imagination. The problem is that people are familiar with their tools and understandably don't want to change those tools. It might be feasible to convince someone to change their email client from Outlook to Evolution and their browser from IE to Firefox but to tell them to change their spreadsheet software from Excel to Open Office would be deemed ridiculous. Excel (in their opinion) displays all their current company spreadsheets with VB macros and tables correctly and a change to Open Office will more than likely garble any non-standard spreadsheets including those which will continue to be sent to them by friends family and colleagues. Equally ridiculous is telling someone who spent four years at University obtaining a degree using Photoshop to ‘switch’ to The Gimp which is 'similar'. Someone needs to raise the bar and sink some cash in to Wine so that MS Office (All versions) and Photoshop (All versions) will run on Linux as if they were native Linux apps. Until these issues are addressed no distribution of Linux will ever threaten the big boys in the Desktop arena.
This may seem like a pretty damning outlook on the state of Ubuntu and indeed Desktop Linux itself but if Linux is going to fulfil its potential it's time we took a realistic look at what Linux is and what it isn’t. This needs to be followed up by developments that will change it for the better rather than just proclaiming it to be the best and waiting for it to 'take off'. Then and only then will it live up to the hype that Ubuntu is currently creating.