Off the Beat: Bruce Byfield's Blog

For nine years, my default desktop was GNOME. About the third of the time, I'd use another desktop or a shell, either for the purposes of review or just for a change, but I'd always return to GNOME. It was a no-fuss interface in which I could do my common tasks without any problem. But a glitch on my system that left GNOME unstartable coincided with the release of KDE 4.2, and -- not having the time to reinstall -- I switched to KDE. I haven't looked back since.

Nobody could have been more surprised than I was. I'd worked in KDE 3.x many times, of course, but I was never comfortable in it. The defaults themes and icons looked so blocky and childish that it didn't look in the least modern. It was so different from GNOME that I might as well have been in another operating system.

So why did I switch permanently? Two main reasons come to mind: KDE's design philosophy and its ability to innovate without dictating.

Minimalism Vs. completism

To start with, I started noticing a difference in design philosophy between GNOME and KDE. GNOME's Human Interface Guidelines advocate a minimalist design in which only the most basic functions are available in the interface. By contrast, although I don't believe that KDE has ever formally expressed its design preferences, you only have to look at apps like Amarok, DigiKam, K3B, KMail, or Marble to see a completist philosophy, rather like that of all those Victorians who set out to write the definitive study of their subject. If a feature has even the remotest connection to the core function, then sooner or later a KDE app will add it.

KDE's completist philosophy creates organizational problems that, in some apps or menus are unresolved. At times, too, it seems like overkill to start DigiKam when you want to do some simple editing, or K3B when you are doing a straightforward burn. If you agree with GNOME's philosophy, you might justifiably call these apps bloated.

However, I'm the sort of user who finds an overflowing toolbox more of a comfort than an intimidation. Eighty percent of my work might be done with twenty percent of the features, but I appreciate that the other features are there if I need them. The same appreciation underlies my preference for LibreOffice over AbiWord, although I recognize that AbiWord is useful for some things, such as quickly opening a Microsoft Office attachment and rendering it with reasonable accuracy.

The point is, KDE is developing for all levels of users. Too often, GNOME -- whether you're talking about GNOME 2 or 3, or Unity -- feels like it is taking functionality away from more experienced users in the name of helping beginning users ramp up faster. That's a worthy goal, but when accomplished through minimalist interfaces, it usually means that new users gain basic competence quickly, then fail to learn anything more. If I'm typical (and, anecdotally, I seem to be), they are probably not even aware that any possibilities beyond the basic exist.

By contrast, when an interface includes more advanced features, I have a chance to expand my expertise. Often, I can find simple ways of doing a task instead of inventing torturous and inexact work-arounds. I am far less likely, too, to have to spend time searching for another tool to do what I want. To have these advantages, I willingly endure the odd patch of disorganization.

A different view of the KDE 4 series

Another reason I switched to KDE was the spirit of innovation that I saw in the early releases in the fourth series. Unlike many people, I knew KDE 4.0 wasn't ready for everyday use, but it was also the first desktop for years that had intrigued me. Suddenly, a group of developers seemed to have realized that they no longer needed to worry about catching up with proprietary desktops, but could finally take the lead.

Never having been a fan of the KDE 3 series, I approached the KDE 4 series with fewer preconceptions than most people. If anything, I was all the more intrigued because nothing in KDE's past had prepared me for the fourth series.

Specifically, what interested me was that -- contrary to many people's hurried judgments -- the fourth series was not abandoning the features of the third series so much as rearranging them to make innovation easier.

For example, instead of being confined to a single desktop of icons, KDE 4.x allows multiple icon sets that can be easily switched. Instead of a single layout, it offers multiple views of widgets and icons. Instead of a single desktop, it encourages the use of several virtual ones. No one needs to use all these features; if you choose, you can ignore them and work in much the same way as you did in KDE 3.x. But if you do use them, you soon understand that KDE 4.x is not breaking with the traditional ideas about the desktop so much as expanding them.

Wherever KDE's fourth series innovates, it seems careful to allow alternatives. Don't like the menu? Then you can not only choose a classic accordion-style menu, but another menu called Lancelot. Prefer the traditional tree view for system settings instead of groups of icons? Then you can have it.

This approach is in marked contrast to both GNOME 3.x and Unity, both of which are full of marked breaks with the past and give you only the choice of enduring the breaks or else finding another interface. Alone among the major free desktops, KDE is innovating without insisting on dragging you along every step of the way. Although some people paint KDE as unwilling to consider the needs of users, the truth is that is actually far more tolerant of different ways of working than most of the alternatives.

Choosing KDE

By expressing the reasons for my preferences, I'm not trying to revive the old flame wars. GNOME 3.x and Unity have their supporters, and I respect their right to choose even while I disagree with their choices.
Still, if you are trying to settle on a desktop, then perhaps my points are worth considering. KDE as a whole has faults, and it has made some mis-steps. Yet, all the same, it is the only major desktop that does not try to force me into the work flow its designers think is best for me (as though interface design wasn't dependent on your starting assumptions). Instead, it offers me the chance to learn more about the software I'm using, and lets me work at least close to the way that I prefer.
All I can say is, it works for me.

Comments