wyoGuide, another way to improve the Linux desktop

Preface

There are currently some initiatives to improve the Linux desktop, to make the Linux desktop better suitable for the average user. The Tango project tries to improve the look of the desktop by defining style guidelines and naming conventions of graphic elements (icons, etc). The Free Standards Group created the "LSB Desktop Project" to improve the binary compatibility between different distributions. There is some on going work at the freedesktop.org which improves various different aspects of the desktop by developing standards. And finally the OSDL set up the Portland initiative which should take care of the Linux desktop itself.

But all these initiatives lack a rather large part of necessary improvement. Because from the user's point of view, the usability of a system is only partially defined by the usability of the desktop. A large part (IMO the larger part) is defined by the usability of the used applications. Usually when a user speaks about the usability of a desktop he most of the time means usability of the applications. Logically to improve the Linux desktop, the usability of the applications has to be improved since usability is the most important factor of the attractiveness of a system.

The wish for improving the application available on the Linux desktop doesn't mean enhancing their functionality. Most applications already are rather high sophisticated, sometimes even more than commercial applications, but an average user has problems using them since he doesn't find the correct commands, doesn't know how to do something to get his work done. By wishing for improving the applications, it's meant to improve the usability, how the functionality of the applications can be found and used by an average user.

The usability of a certain application can't be determined without also looking at other applications. Most applications have a common set of the same or similar functions and all these functions should be done in the same or a similar way. On the other side different functions have to be done in different ways. The usability of an application can only be fully determined in the context of the other uses applications.

OSDL Linux desktop survey

The just released Desktop Linux Client Survey from OSDL lists "application support" as the first top inhibitors of a Linux desktop adoption. "Application support" in this survey means the users wished applications which aren't available on Linux albeit a majority had Linux deployed and probably knew all the applications which are available on Linux desktop.

The survey doesn't answer anything why users wished none-free applications. It also doesn't answer if these users are willing to pay for these applications. It only shows that users didn't wish to use the free available applications. While the missing functionality most probably can't be the reason the only sensible alternate explanation is the users aren't happy with free applications.

LinuxQuestions.org thread

The thread at LinuxQuestions.org about "What programs would you like to see ported to Linux" shows the same problematic again. All kind of none-free applications are wished and just very few alternatives are listed. You can assume that only Linux fans visit LinuxQuestions.org so this thread is even more amazing. So even if all these users have Linux deployed they don't wish the free applications. Poking through the thread gives the impression that users seem to have difficulties to become familiar with any of the free application.

Novell: Cool-Solutions web site

Novell is sponsoring quite a lot of Linux efforts but it seems they have come to the conclusion that not everything goes as expected. They have put up the Cool-Solutions web site which clearly shows that users prefer none-free applications again. I guess many of these users (likely business users) would pay for these applications and just discard any free alternative. Unfortunately also the Cool-Solutions site doesn't provide any reasons for these user preferences.

Reasoning regarding a usability problem

If you use a Linux desktop system you most probably use all kind of applications. Some of these follow the guidelines from Gnome, some those from KDE, some those from Java and a rather large part follow any other guidelines or none at all. So while on other platform all applications look and behave rather similar, on Linux each application looks and behaves quite different. Therefore it's almost impossible to conclude from how one application behaves, how others will. No wonder many users have problems to become familiar.

Just look at a sample issue, how free applications present their help to the user in case one doesn't know any further. Some applications have man pages, some have info pages, other have just some local documentation, some have an online help (hopefully the web is working) and just a few have a context sensitive help. It's obvious that such a help situation isn't very much liked by the users.

What about none-Linux user liking to switch?

While many application do run under e.g. Windows very few do in a platform friendly manner. So all sort of problems arise some might be real some just imaginary. Most important these applications just look and behave different than the other applications. This is worse since these users are mostly accustomed with a rather consistent look and feel, much more than the Linux users. Therefore they expect more and are faster disappointed when it isn't the case. So many potential users liking to switch are just scared away.

How can this situation be improved?

Well the Linux desktop community has to come to terms and finally agree on a single set of guidelines which have to be established. This is a rather tricky task since it seems neither OpenSource developers like to follow any guidelines nor do Gnome and KDE like other guidelines but their own. But it's no question, any Linux user would greatly gain from such a single set of guidelines as would the whole Linux community itself.

Free applications should also compete on other platforms against their commercial counterparts just to show what they are capable of. But they have to look as native applications as true cross-platform applications. Only then users will try out these applications and hopefully become familiar with.

The most important step towards reaching this goal is to establish wyoGuide as this single set of guidelines, so finally the free applications become more user-friendly and improve the tiring experience of the Linux desktop. This also allows every free application to become cross-platform, lifting the monopolists grip on the desktop quite considerable (see Cross-platform, the freedom to choose any computer) and hopefully fix the Ubuntu Bug #1 at least partially. Sure wyoGuide is still work in progress and it really needs sample code for more frameworks, so there is still much room for improvements.

If you're really interested in the success of the Linux desktop it's your task to push wyoGuide and the cross-platform movement forwards. Only then you'll get nice and usable free applications, only then will vendors be inclined to port their own applications to Linux. Everything lies in your hand, you can make it happen.

For more information look at the cross-platform wiki http://wyoguide.sf.net/wiki and see what you can do or go to the developer section http://wyoguide.sf.net/wiki/index.php/Developers.

Conclusion

To improve the Linux desktop it's essential to improve the applications as well. While a Linux desktop is possible to some sort of extend, the final breakthrough will only come when the available applications are used and wished by the majority of the users. Only then the overall picture of a Linux system will fit in the eyes of the users.

To make your application usable you have to follow some sort of guidelines, to make it usable in a greater context you have to follow guidelines in a cross-platform fashion. wyoGuide gives you the necessary help you need to achieve this goal.

To make the Linux desktop a success it needs the participation of everybody. It needs developers who do the actual code writing and it also needs the users who provide the feedback and takes care of any none coding task. Only when everybody is working together will the Linux desktop finally become a success.



wyoGuide Home



Submitted by Otto Wyss, Feb. 14th 2006 - Rights Holder