Update: New, low effort sample pix added!
Please don't get me wrong, It's not my intention to bad-mouth your idea, I'm just trying to help improve it. The validity of my previous comments remains. In addition, with your pictures I can see some more problems:
What will you do if people want
more than 5 desktops? At one point, having a single button with text on it for every desktop will get pretty messy and confusing when having more than 5 desktops, no matter how you display the buttons. On the other hand, when having <= 5 desktops, it always only takes a maximum of 2 "flicks" to reach any desktop, whereas by using the desktop bar you need 2 clicks (one on the Pandora button, one on the desktop bar) and need to refocus your concentration in the process (as soon as the desktop bar shows up), and if you want to close the desktop bar after changing to another desktop then you'd even have to do 3 clicks and refocus your concentration once. So I'd say that the current design of the desktop bar is improvable. What about putting it at the top and showing it all the time, with maximized windows automatically covering it, and using symbols instead of text? By separating these symbols in groups of four, an overview would remain, since 4 is the maximum number of elements a human brain can count instantly without thinking and looking around. Like so:
Code:
-- Desktop bar (every "O" is another desktop)
|
V
|----------------------------------------------------------------------|
| O O O O | O O O O | O O O O | O O O O |
|----------------------------------------------------------------------|
| | V | |
| --------------------- |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
|----------------------------------------------------------------------|
| P --------- ---------- --------- - - 12:34 -|
|----------------------------------------------------------------------|
The symbols could be micro-pictograms indicating the location of items on the respective desktop, that way their look would be coherent which would avoid confusion while making the desktops more identifiable through their symbols.
When the
drawer is opened and the sorting buttons appear (by the way, it's much better to use "Recent" instead of "Date", since "Date" is frequently misinterpreted as "Date installed" or something else), what will you do if people want more than 5 buttons (which they most probably will, since the way of sorting is also a personal preference)? Making the user scroll them left/right or up/down would be confusing and more often than not need more effort than just scrolling through the application icons. Stacking them in several rows would take up valuable space that could otherwise be used for displaying more application icons at once. So there's room for improvement too. The first idea that comes to my mind here is a rainbow color system where the applications' icons are sorted by their dominating color. Some of the (all except for one?) sorting buttons are replaced by a rainbow slider, and by touching the slider/scrolling on it using an analog nub, the application list will scroll to the selected color. That'd have some accessibility issues though, so it might be good to build in an alternative, like for instance a search box that automatically opens up as a button on the keyboard gets pushed when the drawer is opened.
------------------------------------------------------------------------
For minimenu I've built it so 'touch to run' or 'touch to change to that tab', but I've not built in gestures or anything like that .. "swipe to change to next tab". If someone wants to make a good case for gestures -- that people are interested in that metaphor of operation, it might be worth pursuing.
Just that people are interested in something doesn't mean that it will turn out to be useful. Personally, I pay a lot of attention to communicating with users who suggest a feature. Very often suggestions turn out to be a case of
argumentum ad populum, in which case implementing them more often than not would be a particularly bad thing.
Also,
by copying something because it's "hot" (~="has turned out to be very useful in some other projects") without thinking twice about its actual usefulness, more often than not a software can become good, but probably not excellent.
Anyway, not having a Pandora I can't really say anything about how useful or not gestures would be in minimenu. Just be aware of the dangers of "sparklingarbage features".
------------------------------------------------------------------------
edit-08jan2011-05:39GMT+1:
[*]Some minor changes
