Flip3D: New != Improved
March 20, 2007
An important facet of knowledge: knowing what you know, knowing what you don’t know, and knowing what others know and don’t know.
I don’t know Usability the way I should, but I know that certain engineers at Microsoft are not in much better shape. Witness the monstrosity that is Flip3D. Everyone raise your hands: Who here winced the very instant they saw the implementation of this application switching disaster?
The series of missteps required to research a solution, code it, test it in user labs, and get the greenlight for production are mind-boggling. Right now, the Pontiac design team is breathing a sigh of relief, as their Aztec looks downright competent and slick in comparison.
Can you imagine the whiteboard at the first design meeting? It only had three points:
- Must be animated
- Must not look like Expose
- Must be done by Friday
They’re breaking Wade’s Cardinal Rule: Can Do vs. Should Do. Why are the windows placed in that direction of back-left to front right (are they reusing the card-springing code from their Solitaire app)? Why are they placed in depth perspective? Why are we inducing the need for antialiasing on everything and begging for jagged pixel lines because not a single line seems to be horizontal or vertical? Why do users want to see distorted angular versions of their miniaturized windows? Why do we think users only want to see the front window in its entirety? Why have we come up with a solution that will make a user’s brain explode if we ever expand Flip3D to cover virtual desktops? Why have we created a solution that avoids any sensible future ability to cursor drag and select multiple windows?
Why, because we can! Because we’ve got to show that we can do desktop acceleration. And we only had 3 days.
More reason for the KDE community to be thankful for our Usability experts. If you know that you don’t know usability, know that there are people who do know it and know they know it.