Comment: Unity can't go it alone
by Thorsten Leemhuis
Open source experts like Red Hat and SUSE learned long ago how difficult and often unwise it can be to try to establish bigger projects with just their own distributions. If Canonical realises this, too, and changes its methods of operation accordingly, the Unity desktop will have a much better chance of becoming the third major desktop alongside GNOME and KDE.
Although well established projects and companies working with Linux certainly have their own influence, most open source software has only spread far and wide because it was able to convince many users on its own merits. Only then are people interested enough to suggest improvements, report bugs, and test release candidates; only then do users blog about tips and help new users enough that, in the end, the program greatly benefits from a large ecosystem.
It is therefore daring – perhaps even foolish – for a distribution to try to make a large software program successful completely by itself. But that is exactly what Canonical has been attempting for about a year now by using Ubuntu's popularity to establish its own Unity. Because of a few hitches and quirks, the desktop environment is an isolated application that has hardly any chance outside of the Ubuntu universe.
Indeed, Unity can only live up to its full potential if some of the components already included in Linux distributions also include extensions developed for Unity. These changes are mostly needed for software for the GNOME Project and its environment, such as GTK+. The Ubuntu packages for GTK+ include the Unity extensions, but most other distributions scorn them. These problems have been growing for some time now, and a year ago they led to the end of some efforts to offer Unity for OpenSUSE and Fedora.
New efforts have since begun or are being discussed; repositories that provide Unity for Arch Linux and openSUSE are working around the problem by offering their own packages with GTK+ and other programs instead of the packages belonging to the distributions. That may sound harmless, but it is just the beginning of major DLL problems, which would present themselves as soon as you, say, want to use another program that is dependent on other GTK+ extensions but doesn't include extensions for Unity.
Canonical could have – and should have – avoided this problem by giving its developers more time and resources to include the Unity extensions for GTK+ and other programs in the upstream projects. Indeed, they certainly tried, but were only partially successful. It would be too easy to place the blame entirely on those maintaining GTK+ and other programs – even though it's very tempting, since many of them not only work on Unity's "rival" GNOME, but are also employed by Canonical's "competitors". But those who personally know developers who have been active in the open source world for some time also know that the aforementioned aspects are rarely the actual cause of a problem; ego, misunderstandings, stress, technical problems and the pursuit of a technically better solution tend to be much more common explanations for why some developers disregard improvements.
Most of us can think of similar issues with colleagues in our daily work life. And just like in the working world, in the open source world there are also possibilities for getting around people and superiors who you don't like, or who are resistant, so that you can still achieve your goal despite obstacles. Only if Canonical manages to do so can the Unity desktop be included in other distributions' standard configurations and therefore have a real chance at becoming the third major player alongside GNOME and KDE. The makers of Mint have already realised that they can't go it alone: their recently released Cinnamon 1.2 desktop interface can also be used with other distributions without major problems.