Linked by Thom Holwerda on Tue 27th Jan 2009 13:46 UTC
Editorial With Windows 7 having made its grand debut, and with KDE4's vision making leaps and bounds forward with every release, we have two major software projects that have decided to implement some fairly drastic interface changes. Such changes are bound to receive some harsh criticisms - but the funny thing is, these criticisms usually come from people you least expect it from.
Permalink for comment 345798
To read all comments associated with this story, please click here.
80/20 Rule
by elsewhere on Tue 27th Jan 2009 18:47 UTC
elsewhere
Member since:
2005-07-13

The common thread in all of the criticism about KDE 4.x and missing functionality is that the complainers dismiss the project as a failure because it doesn't meet their personal requirements. Any project that strives for 100% user satisfaction is doomed to stagnate from a lack of innovation due to developer fear and paralysis of alienating any potential user.

I'd argue that 80% or more of KDE's legacy featureset was utilized by 20% or less of the user base. When decisions need to be made as to prioritizing development objectives, the project can:

a) strive to maintain functionality for the core user base at the possible expense of the "poweruser" base

b) strive to maintain functionality for the powerusers at the possible detriment of development for the core user base.

c) strive to maintain all existing functionality while somehow managing to build new features and create a forward looking foundation

Certainly many would say that c) is the ideal to strive for. But I see two funamaental problems with c). First, the KDE developer pool is limited and that is the primary reason that priorities need to be set and sacrifices need to be made, even if only temporarily. If KDE waited until 4.0 was ready to meet everybody's expectations, then it would still be stuck in a never-ending development cycle, and 3.5 would stagnate and suffer from lack of developer attention.

The other issue is the idea of ensuring legacy support to avoid alienating any of your existing users. This is the MS model that insistst that every version of Windows should be as backward-compatible, which introduces a whole different set of issues. Security issues and fundamental flaws from the legacy platform are inevitably incorporated into the "new and improved" platform, and it just becomes a mess, even if the user experience is preserved.

Many of the features that are missing from the KDE4 transition were hacks upon hacks built upon the legacy codebase. They were often unmaintained and undocumented. Trying to migrate them into the new codebase could have caused additional issues with support, stability and code development.

You can't make an omelette without breaking a few eggs. These examples are becoming cliched for having been mentioned so many times, but looking back at OSX 10.0, Windows XP, Gnome 2.0, Kernel 2.6, KDE 2.0 etc., all of those projects met the same amount of derision and criticism. All of those projects evolved into stable, full-featured systems that many users now hold as the paradigms for platform stability and capability.

The problem with the 80/20 rule is that the 20% complaining gives the impression that the other 80% are "suffering" too. People complain far more often than they praise, and often far more loudly.

In my own un-empirical experience, I've demonstrated KDE 4.2 to non-tech savvy users in my office. People that have never used linux, or have barely touched it. Simply showing them the start menu was enough to for them to start web browsing, launch a word processor etc. Even accessing shared drives on our MS network. In some cases they even commented that Dolphin was preferable to Windows Explorer. Their impression was that the desktop was polished and "professional" looking (I keep my desktop effects moderate and don't overdo them), and surprised those who were at least familiar with linux in as much as knowing it exists.

To me, that is important progress for the KDE project. That is the direction they need to move in. KDE cannot hamstring itself by being a poweruser desktop, it needs to embrace a wider user base. The poweruser features can be added in a sane and rational manner as the DE evolves, but their narrow usage requirements shouldn't drive the entire scope of the project.

TBH, I agree with the general point that the non-savvy users are actually much more willing to accept change. Yes, they may rely on knowing exactly what folder something is in, or the steps to do something via menu features, but by that same token, when you direct them to do something differently, they simply adjust their process. There's a learning curve, but I think the size of that curve is vastly exaggerated. And perhaps more pointedly, they don't start blogging about how awful it all is.

The KDE team moved the cheese. Some are adapting, some are stubbornly refusing. There's nothing different there that couldn't be applied to any other project.

It is sad when a large-scale OSS project loses users, but it's to be expected, and can be vastly outweighed by the potential to bring in new users.

Besides, I suspect most of the "switchers" will be back soon enough. Complaining about missing functionality in KDE and switching to a certain alternate DE instead has a bit of a hollow ring to it. KDE 3.5 is still around and will remain around for some time, giving people a chance to transition when they're ready.

Just my 20c...

Reply Score: 8