UProtokoll 06 04 05
Open Software Usability Stammtisch
Protokoll vom 5. April
(von Ellen Reitmayr - heute in Englisch, da das Protokoll auch an die KDE-HCI-Liste geschickt wird)
In a session about the 'vision' of KDE, three members of the KDE HCI working group and some "neutral" members of the Stammtisch analysed statements and comments from different "key people" and movements in KDE, for example:
- KDE 3.5 weaknesses from hci pov (see hci mail archive)
- HCI Goals (see hci mail archive)
The goal of this analysis was to learn what KDE really wants, and to formulate a vision from it. We took a step back and had a look at what other people in KDE think about the project's future - trying to function as a catalyser, extracting long-term goals from others' statements, and looked at our own HCI goals in that light.
While we extracted quite solid statements regarding the style and interaction design of the future KDE desktop, there were only vague statements regarding the target users.
"Developer", "Geek", "The common information worker", "The mainstream user", "Grandma", "The Chinese rice farmer" 
Appeal also says:
"KDE desktop systems are started to be rolled out in enterprises, offices and organizations, sometimes even in large numbers. We begin to take pockets in places where proprietary systems where 100% predominant previously. But still, our competitive strength is not yet enough for "world domination". 
this may be interpreted in a way that we follow a strategy to conquer the home user desktop via enterprises, and that we are heading to gain 90% of the desktop market.
"Looks aside, Plasma must offer the average KDE user efficient and ergonomic access to their desktop first and foremost." 
This definition is very fuzzy: The average of the current KDE user base, of the one 2 years, in 5, or even of 5 years ago?
While the one of appeal was quite definite, we hesitate to take over its definition as the project somehow ended in talk. It's beyound our scope to say if it was due to the goals Appeal set, due to the rather business-oriented strategy, or due to other reasons (choice of members, momentum, ...).
Why is a proper definition of target users so important?
Without a definition of the target users, all our vision statements remain blank ideas. We cannot design an easy-to-use desktop if we don't know what our target users understand and need. Only if we imagine a real user sitting in front of our desktop, working on specific tasks, we can build a vision of what the software needs to look like in order to support him best.
In our session, we found that it cannot be the HCI group who decides about the target users, but that we need to analyse the wishes in the project, especially the ideas of key people in the project - people, whose opinion is respected and followed by the other members.
From a HCI pov, the design of an optimised desktop is easier the narrower the target user group. We know that clashes with other opinions in KDE, that's why we won't promote it without hearing the others', well-reasoned wishes.
Peter suggested to use the mid section method for that:
Do you want more than half of the desktop users?
And in now, do you want more or less than half?
And this time?
Vision Statements so far
The first thing we learned: Visions are the top of the mountain. They may sound unreachable at first, but they provide goals and directions and allow the formulation of subgoals and milestones. It's better to "think big" in the first run, than to realise later that you forgot about important things and can't house them anymore.
Apart from the missing and essential definition of target users, the following vision statments were extracted from the different documents:
- KDE is workflow-oriented rather than application-oriented
The desktop supports the user's workflows in an optimal way and provides functionality on demand. The desktop is the major metaphor that provides and integrates this functionality. The whole desktop is perceived as a wholeness, this includes a consistent look, feel and interaction design.
- KDE supports relations rather than physical position on the hard disk
Information is presented according to relations, not to the position of the hard disk. Primary concepts are mapped in an easy way.
- KDE is easy-to-learn
The initial understanding of the desktop is granted, that means only the primary concepts are presented in the first run (!! this absolutely requires a definition of the target users!!). functional defaults are given, but options can be extended on demand. a beautiful and intuitive design supports the understanding of options and relations.
- Accessibility is a major goal of KDE
- KDE is stable as a rock
Extensive bug testing, may include usability "bugs"