__/ [TheLetterK] on Sunday 12 February 2006 20:19 \__
> Roy Schestowitz wrote:
>> __/ [John Bailo] on Sunday 12 February 2006 17:52 \__
>>> Roy Schestowitz wrote:
>>>> "...But the biggest surprise is the new graphic interface which uses
>>>> OpenGL to create effects which you are all too familiar with, like
>>>> window transparency, switching desktops using a large rotating cube, or
>>>> creating a thumbnail view of a window."
>>> The thing is, people have been "featuring" crap like that for years.
>>> As to whether it enhances the GUI experience, or information delivery,
>>> or anything useful -- is really unknown.
>>> My opinion: A well designed 2-D interface is far more useful.
>> I agree. As long as a third domension is not supported by interaction
>> peripherals (4 arrows on keyboard, 2-D mouse motion), that third dimension
>> is an overinflation that cannot be used.
>> Now, imagine a mouse that had a notion of /depth/. I saw one on TV when I
>> was about 12. You could then grab a window that is hidden beneath another
>> windor that shows up on front. You could trivially push and pull windows.
>> If an O/S vendor took forward this paradigm of HCI and nurtured it, that
>> would become a huge selling point.
> I know how you could do it with existing mice. Use the mouse wheel for
> traversing the z-axis. Not hard at all. Certainly less cumbersome than
> the UI shown in that video. Alternately, you could start using the
> keyboard for level navigation. Or you could use a gyroscopic mouse, that
> detects z-axis motion (these already exist).
A gyroscopic mouse is what I was reffering to all along. Anything else is an
emulation of the z-axis, which amateur users will find counter-intuitive.
One flaw comes to mind, still. With 3-D environments and 3-D mice, you would
still need good 3-D diplays or exploitation of stereo vision, at one level
I suppose you could use mouse pointer size to indicate depth, but it's far,
far from ideal. Meanwhile, one could think of virtual desktop as a /stack/
of desktops, lying one above the other, much like layers. It gives you
pseudo-3D. The convention, however, is to treat them similarly to the way
they are stored in a memory frame or (virtual) frame buffer. We think of
them as one extended rectangle, which probably simplifies matters.
All in all, I think that a sphere (of let us say 6 desktops) would only
confuse users. Think of the globe vs. map analogy. One day our children will
laugh at us for thinking that the desktop was bound to be flat.
>> <snip />
>> <snip />