gnome

Since I became a regular Linux user around 2003 I was a KDE user. I stuck with it on half a dozen distros along the way. On a whim, two weeks ago, I switched to Gnome and haven't been sorry, yet. This post specifically focuses on KDE within Kubuntu, compared to Gnome in Ubuntu.

Many of these are personal observations which have lead to my switch. It's a rant at times but I want to say very clearly that these comments concern the software and I appreciate the effort of all the volunteers on KDE and Kubuntu very much.

Why choose?

Most people from the community who I have met tend to prefer one desktop environment (DE) over another, or at least choose one and mostly stick with it. There are a plethora of choices when it comes to window managers, file managers and so on. Personally, I could never really appreciate the more bare-bones ones, such as Fluxbox, simply because it soon became too annoying to set up every little detail by hand.

More importantly, though, the underlying libraries in the DE often add additional benefits. For example, in KDE your address book could be accessed by the mail and IM application. Throw in a Gnome application and it has no clue about the address book, at least not without customization.

Many things have improved over the years, such as a common user notification method. The user, however, can often still assess quickly that one app simply does not fit when it is placed next to apps from another DE with different HIG. It's probably superfluous to mention that Apple excels so very well in this department.

Switching costs

When you consider moving from one of the main DEs to another, there are some opportunity costs. You will have to relearn ingrained tasks and shortcuts (e.g., hitting ALT+. reflexively in the Terminal window and nothing happens, is annoying) and finding appropriate replacements for each application. Nonetheless, if the perceived benefit from switching exceeds those costs, it should be worth it.

In my case, it was clearly worth it. From earlier short stints into Gnome-land I had the impression that I'd probably have to keep several KDE applications because I would be missing key functionality but it turned out that this was not the case.

Issues

It is difficult to disentangle Kubuntu from KDE here to see where the problems and thus reasons to switch originate. It seems to be a combination of several factors:

Community distribution:

It's plainly visible that the primary support to make Ubuntu a great distribution goes into the Gnome part. Kubuntu often has to adjust to architectural changes for a release, whether or not KDE upstream has caught up to those as well. This makes Kubuntu work less seamlessly. Examples: Pulseaudio, NetworkManager, Bluetooth, etc. It's probably also the toughest problem since Kubuntu as a distribution could probably not function, when detached from the rigid six month release cycle of Ubuntu. Apart from using a completely different distribution, such as Mandriva, I don't see a realistic solution to this.

Kubuntu's & KDE's priorities:

When KDE 4.0 came out I was excited and switched to it and stayed until 4.4.2. Initiatives like Plasma and Nepomuk are cool ideas. However, it really doesn't matter so much that the weather widget is pretty and the network widget is shareable if you can't get on the VPN without using vpnc & openvpn directly. Or if attaching an external monitor hard-locks Xorg (but Gnome has no problem), or half a dozen other issues every single day. 28 months after the official release of KDE 4.0 all these minor issues should have become a priority to make KDE a reliable DE in current distributions.

KDE's planning:

While there were release schedules and clearly plans for changing underlying technology such as Akonadi over the individual releases, a clear outcome seems to elude them. Ignoring for now the three iterations of indexing backends we went through with Nepomuk/Strigi, it's just not OK to have two embedded MySQL instances running solely to write an email and listen to music. Especially if upgrades result in core parts, such as Akonadi, crashing on each first start, no matter how many tutorials you try out.

In the future

Today, from my point of view, KDE is overbuilt and lacks stability as well as too many critical features for me to work effectively. I'd certainly consider going back at some point. I don't think this will be any time soon. Especially since the benefit from KDE's advances would also have to go beyond the switching costs, again.

Nonetheless, I'd like nothing more than to be proven wrong.