Re: [Usability] Tab implementation review
- From: Diego Moya <turingt gmail com>
- To: usability gnome org
- Subject: Re: [Usability] Tab implementation review
- Date: Wed, 15 Jul 2009 12:48:16 +0200
2009/2/25 Philip Ganchev wrote:
> "What if you want three levels" of organization (tabs, windows,
> workspaces)? Well, I tend to think that having more levels complicates
> the mental model and also requires more time for habituation.
It can be done without those problems, though. Have you seen the
interface in MS Office's OneNote? It has a hierarchy of three levels
of tabs. Those on the left are for notebooks - akin to documents. Each
notebook contains several sections (tabs on top), conceptually similar
to tabs in a spreadsheet. Tabs on the right are "pages" inside the
current section.
This model is simple to understand and quick to learn, since each
level has a corresponding metaphor. And it provides a really good
navigation structure, much better than a flat list (which requires you
to scan the whole list or resort to keyword search) or a full-blown
recursive tree (which doesn't offer any distinction between different
levels).
So there *are* some advantages to a controlled hierarchical
classification. For me, right now, a windowing system that didn't
allow for window grouping and/or placing tabs inside a group, would
feel broken.
As for the solution given by Steve (using virtual desktops as the top
hierarchical level, and window lists to retrieve open documents) it
has a severe drawback with respect to tabs, which is that of
visibility.
I never use virtual desktops because, for me, a window in another
workspace is "lost". I rely on the persistent visual cues of open
windows to remind me of ongoing tasks. Tabs enhance this for
particular documents; it allows me to juxtapose the tab list in
several windows at once, giving me the chance to quickly find where a
recent document is placed and to which task it does belong. All this
is lost with a single unified list of everything, or with tasks
distributed among several desktops.
As for the other reasons stated, none of this is relevant about a tab
interface *per se*. They exist because current implementations of tabs
only allow to share tabs from the same application. A tabbed interface
allowing to group tabs from different applications would allow placing
all documents related to a single task together, while keeping all the
spatial advantages that I stated above.
[Date Prev][
Date Next] [Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]