Hi, I'm new to the list. I love how KDE developers in general are embracing a usability culture, and I'd like to be involved in design of the "interaction architecture" for the 4.0 release. Though I feel that in the Open Source world the academic research on usability is not known enough, and several myths on users forged on the "developers" front are repeated once and again. I hope that we developers learn to hear more from the HCI experts and finally embrace the known design techniques and lore from the cognitive scientific research. On Sat, 12 Mar 2005 02:35:56 +0100, Sven Burmeister wrote: > From my experience there are three kind of users, which have all to be > considered. >First, users that are reluctant to learn/understand/remember [...] > Second, beginners that want to learn/understand/remember. [...] > Third, advanced users. I'm afraid you've fallen into the trap of the "user levels" meme. There are several reasons why this compartment of interfaces into levels of expertise is a bad idea. Right from my mind now I can recall these: * A user may be beginner for some tasks, and expert for others. * It is difficult to say whether a user is novice or expert, even for oneself. * Users learn. When a user has used an application for a while, is at ease with it and want access to the advanced functions, dropping her into a changed interface (the one for the next expertise level) would be a terrible thing to do. * How do you categorize expert users on rarely/seldom used applications? You can't expect them to remember the details of an app used several months ago, but the user want to accomplish expert tasks, not just the basics. (These are called "transient" applications, and there are specific design tips to create them). Here is an article where a much better model for user learning and "intuitive" interfaces is explained: http://www.webpronews.com/webdevelopment/sitedesign/wpn-26-20050117WhatMakesaDesignSeemIntuitive.html It includes the concept of the "Knowledge Gap" to complete a given task in an interface. Where the Gap is wide, the user must be trained before she can complete the task. An intuitive interface would be one in which the training is done in small, discoverable steps. Note that any user might benefit for a discoverable interface for incremental learning, not just office types. There ARE some scenarios where it's better having a training program to form expert users and make them proficient with a difficult but high-performance interface. But this will only happen in corporate environments, where paying for such a program would pay off, and users are highly motivated (to keep their jobs). A top-level heuristic (rule of thumb) to design user interfaces is "treat your user as very clever but very busy". So a good interface should be designed putting all users into your "reluctant to learn/understand/remember" category, not requiring that users "want" to learn your application before they're able to use it. Also note that an expert user with a desire to learn (and not just *use* the app) is most probably a programmer, so she would most benefit of having an API not a GUI. That's another reason why a GUI should treat all users as busy and not get in the way of getting tasks done. Back to the KDE desktop, a way to achieve the "incremental learning" would be presenting at first use a walk-through tutorial on the very basic interaction techniques (right and double clicking, on-mouse-over tooltips, using menus, keyboard accelerators, etc.) and then requiring in the HIG that all applications make their functionality discoverable through these standard techniques. For a user knowing these techniques, any application that makes its features easily findable would be intuitive. Several distros are taking this same approach and provide a "Welcome" screen that shows some tasks to do when you first use the computer, but they usually don't include this basic training - they show the available applications, but don't teach a common ground that all users should know. Microsoft got it right since Windows 98, and I believe that's one of the reasons it's widely considered as easy to use, even though it isn't. Diego.- _______________________________________________ kde-usability mailing list kde-usability@kde.org https://mail.kde.org/mailman/listinfo/kde-usability