Dear Linux Magazine Reader,
Our cover story this month on the competing desktop technologies Xgl and AIGLX highlights yet another example of the rivalries at play in the ever-changing world of open source. But at least these fledglings get along better than many others in the nest. There is something about the GPL that helps rivals get along. Beneath all the complications (some personal and some technical) is the realization that, if your code is better, I'd be crazy not to use it, and if my code is better, you'd be crazy not to use it. Of course, problems always arise around this issue of what you mean by "better."
You'll learn lots more about Xgl and AIGLX later in this issue. For now, I'll return to another rivalry that has been in the news recently. I used this space two months ago to give my opinion on the quality of the rhetoric that has recently passed between the supporters of KDE and Gnome. (To summarize, my position is: everyone should just choose the desktop they like and quit screaming about the desktop they don't like.) Of course, a few loud voices sound like a revolution. The greater portion of Gnome and KDE users share my sense that there is no very good reason for trashing someone else's desktop. Still, since I made a point of singling out the controversy, I should also shine some light on a more positive development.
When I attended the Boston LinuxWorld a year ago, I was struck by the anguish of the OSDL desktop panel as they pondered the very real possibility of having to choose between KDE and Gnome for their desktop reference specification. Of course, they (and everyone else in the room) knew that this problem of writing software for the two different desktops was a huge limitation for Linux, but the idea of choosing one instead of the other seemed very divisive.
When I saw the OSDL desktop panel at this year's LinuxWorld, they were onto something new, and they seemed much more optimistic. As you may have heard in April, an OSDL-sponsored group called the Portland Project announced a set of common interfaces that let the developer write one application that works with Gnome, KDE, and a number of other desktop environments. The original vision was to start with two interfaces. The first would be a set of command line tools, and the other (possibly more influential) interface would be a common API called DAPI that applications could link to as an interface with the desktop.
News sources quote Portland member Waldo Bastian as saying, "there exists a perception that developers need to choose between targeting Gnome or KDE," when "in reality, 95 percent of desktop application functionality is independent from the window manager and desktop environment. It's the 5 percent that the Portland Project is addressing."
If the Portland Project is successful with their plans, all differences in the available desktop tools will be easily resolved. Of course, just because a committee imagines a solution, it doesn't mean the world will suddenly adopt it. Many KDE and Gnome users I know are perfectly happy with what they have, and they won't be in a hurry to invest hundreds of hours in achieving the status of being a user of GNOMEKDE (pronounced Nome-ca-dee or G-nome-ca-dee). But then, if major Linux vendors would just implement this new API, so that users know what they can expect and that they can depend on it, of course they will use it.
Why not?