On 12/05/05, Daniel Borgmann <daniel borgmann gmail com> wrote: > On 5/12/05, Diego Moya (a.k.a. TuringTest) <turingt gmail com> wrote: > > In my perfect interface, commands would be invoked through pie menus > > or gestures, both accessible with the right button. > A downside is, that the user is required to hold down the right > mouse button to select other actions, but I think that's justified > (and the accessibility problems could certainly be overcome). > However, are there any piemenu-like designs which would work well for > our cause? It would have to scale to a reasonable number of actions > and ideally it should not be patented... Take a look at the easyGestures extension for Firefox: http://easygestures.mozdev.org/features.html It can easily accommodate up to 12 actions thanks to its upper expandable submenu, and even 24 actions for the advanced "two pressed buttons" alternate view. I have had problems evangelizing its use, though. The flashy circular appearance is intimidating at first, and the icons-only default interface is no good for learning to use it. A more square-ish "checkers menu" with text labels would probably feel better, being more similar to the classic context menu. See the attached mockup of the result. No patents of this design at all ;-) > Of course we could do half of this without the piemenus already, by > just selecting the first item by default and activating it if the > button is released right away. But I'm afraid that the ergonomic > disadvantages of this would be too severe for primitive list menus. It > would also break if a menu can't appear at the cursor position due to > space constraints. Not if at first it only shows the text of the default action, and expands to the whole menu only when dragging. Maybe that would be less discoverable, but context menus is something that users must learn about anyway. Also invocation of many commands could be done through drag-and-drop, a la Mac Os. Gnome is appropriate for doing that. This way, using the context menu would be less necessary. > > In links, you have the problem of trying to select a word inside it > > (it's impossible), for example if you want to parse it into a > > spellchecker or a dictionary. > > Indeed, this has annoyed me more than once... It is a somewhat > questionable exception (or let's say a corner case). > Yes, it's not that important. I'm more concerned with the collision of gestures for selection and action in application and desktop objects.
Attachment:
square-wheel-menu.png
Description: PNG image
