Experiment with KDE desktop

Although the Linux kernel first emerged in 1992, it was 10 years ago, in 1998, that the operating system really began to develop into something more than a project for enthusiasts.

This was the year that saw some important developments with Linux and free and open-source software; investments from Intel, IBM and Oracle, the first major corporate investments to be seen; the open release of Mozilla, the code for the Netscape web browser, which ultimately led to the creation of Firefox; and the first stable release of the KDE desktop.

At the time, the Linux desktop comprised little more than a few basic windows on a plain graphical display. Menus could be created (often by hand, with complex text configuration files) to launch programs you would typically otherwise run from a terminal window. From such a sparse desktop, the early versions of KDE were certainly playing catch up.

Comments such as “it is almost as good as Windows 95” were not only common, but considered to be a good thing. KDE did, of course, develop substantially over the subsequent years. Technology introduced in KDE 2 (released in late 2000) was substantially better, including the use of a network-transparent input/output architecture (kio-slaves), the Konqueror web browser, and the KOffice suite.

These new features were to be more mature and integrated in the KDE 3 series, initially released in 2002, and still developed and enhanced today. And so, 10 years on, KDE 3.5 has grown to become a stable, integrated and advanced desktop, offering a number of features not found in any other desktop environment.

This progression of development has now ceased, however, to make way for the new, and fundamentally different, KDE 4.

KDE 4
KDE 4.0.0 was released in January 2008. It was a rather controversial release, however, and not one accompanied by the fanfare you might expect from such a major new version. The developers have tried (somewhat in vain) to emphasise the difference between KDE 4 as a project and KDE 4.0.0, the latter being the first ever official release of the new desktop, and not a stable, production-ready desktop.

The developers see KDE 4 being good enough to last the next 10 years, with the new technology improving and developing with each release. We’ll take a look at some of these new underlying features to see some of the things that have changed. KDE 4 also provides a number of frameworks, each given a marketing-friendly term to describe its function.

The biggest underlying change comes from the new version of Trolltech’s Qt toolkit. Qt 4 improves upon efficiency as well as design. Together with the improvements in KDE 4’s architecture, the memory usage of the desktop is reportedly set to drop by as much as 40 per cent over KDE 3.5, which is quite remarkable in itself. Another change comes from the licensing of the libraries.

With earlier releases Qt was available under an open-source licence only on Linux and Unix platforms, which has meant very limited use on Mac OSX and Windows. With this restriction gone, KDE applications can now run on these platforms too and, thanks to the cross-platform nature of KDE 4 and Qt, porting software between platforms is almost trivial.

Solid is the new hardware Application Programming Interface (API). This framework allows KDE applications to get information about the hardware they are running on. Previously the developer would have to understand how the operating system presented this information, which will differ completely between Linux, FreeBSD, Mac OSX and Windows.

By using Solid, a KDE developer need only know the standard API functions KDE provides, and not know about the operating system the application may run on. A CPU-intensive application could, for example, modify its behaviour if it found it was running on a quad-core CPU. A network utility would be able to discover if there was a network connection available or not.

Phonon is the multimedia equivalent to Solid. KDE developers now can play back video and audio by using this simple API. As with Solid, the developer would previously have to understand the multimedia system, and with so many completely different options around, developing multimedia applications was incredibly difficult.