Re: [Usability]Re: Nautilus preferences proposal

[David Moles <david moles vykor com>]

On Wed, 2002-05-01 at 10:21, Calum Benson wrote:
> 
> When the horizontal browser-type interface first appeared on NeXT
> desktops, I seem to remember it was because it had won out quite
> convincingly over tree controls in usability tests... people just seemed
> to 'get it' much more than they did with tree controls, with their
> fiddly hotspots and hidden branches.

I suppose that's possible. I'd rather see some attempts made to solve
some of the tree usability problems -- more visual feedback to make
it clear where the hotspots are, what the current selection is, where
(e.g.) a newly created folder will appear -- and making the newly
created folder appear where the user would expect rather than as a 
new top-level item, that sort of thing. Row shading would be a big
plus, for instance.

I also suspect that the usability tests that produced the win for the
horizontal browser only addressed using it to drill down through a
hierarchy to find a file, which you would then open in some other
application and work on for some time. My intuition is that a tree
browser is more efficient if you have to do a lot of switching 
around among widely separated leaves, and/or if you want the browser
to be persistent on the screen while you're working with the things
being browsed (the horizontal browser takes up much more space to
display the same number of hierarchy levels, though it does show
more information at each level).

But I think you could have a horizontal browser where the columns
were tree views and make both sets of people fairly happy.

I'd also like to know whether the horizontal browser was a clear
usability win over the old-fashioned Mac finder that only had icon
views, but remembered window sizes and positions and icon 
arrangements. I think it would be hard to test that, though, because
some of the benefits and drawbacks would depend on how the browser
design shaped the users' file organization habits over the long term.