On 26 Jun 2003, David Legg wrote: > End of thread. OK, but just one more thing (hey, I'm a glutton for punishment)--- > > > Currently, AFAIK, KStep is the only window deco that obeys the last > > two sub-bullets of the Fitts's Law requirement, but it's unsuitable > > for other reasons (the deco icons are weird, and it has only 1-pixel > > resize handles on the left and right window borders, which I like > > personally but which are not what people are used to). > > Note that the Fitts's Law requirement implies that no window deco with > > curved corners is suitable, unless it captures clicks in the invisible > > area between the curve and the corner. > > > All the "easily reversible" window decorations/actions should probably > > do this, except for the Close button which should *not* do this. > > Mis-clicks on most windows decos is acceptable because you can easily > > reverse what you did, except the case of the Close button. > > Usability is _extremely_ important, but if you start analysing too much > you can get way too far up your own backside and overlook the work that > really matters, as some of the discussions on the Gnome mailing list, > and 'features' that have made their way into Gnome, have proved. The > above illustrates my point. Let's not go there. I know you probably were talking about the presentation of those emails, not the content. However: Fitts's Law is not an opinion, nor is it the product of overanalyzing. It's a very simple, objectively quantifiable user interface principle: http://ei.cs.vt.edu/~cs5724/g1/ Fitts's equation does not yield accurate results for 2-dimensional movements, but the principle that larger targets are easier to hit remains the same. It's even been incorporated into the KDE UI guidelines: http://developer.kde.org/documentation/design/ui/fittslaw.html The facts that the Keramik window deco + does not fully exploit screen edges and corners + has relatively small window action buttons are, objectively speaking, usability flaws. Or, at least, engineering tradeoffs made in the name of some other goal---visual distinctiveness? Whatever. You can make up your own mind. However, these particular flaws seem fixable without tossing the Keramik deco entirely. ~k -- GPG public key id: 0x5CFD1761 (available on a key server near you) _______________________________________________ kde-usability mailing list kde-usability@mail.kde.org http://mail.kde.org/mailman/listinfo/kde-usability