Re: Nautilus splitted-view

[Cody Russell <bratsche gnome org>]

On Wed, 2008-02-13 at 23:39 +0400, Christian Kirbach wrote:
> > Isn't window placement/management job of window manager, not
> Nautilus?
> 
> That is not what Gilles was referring to. For a drag-and-drop
> copy/move in  
> a situation
> similar to the initial screen shot with two nautilus windows, you
> first  
> would have to move
> the two windows side by side. Metacity won't do that for you
> neccessarily.
> If there was an option "split view" you'd be one click away from the  
> window arrangement
> depicted in the screenshot. Hence, saves some window shifting and
> resizing.
> I believe this is the initial idea of the thread starter.

It still seems, to me anyway, like this pollutes the UI in an effort to
work around a deficiency in window management.

The pollution is in particular the things that Tuomas mentioned:
  - Multiple views require multiple location bars
  - Multiple views require multiple forward/back/up/home/reload/etc

What do you do with forward/backward/up?  Do you have multiple sets of
them?  Or do you have one set that applies to whichever pane has focus?
Then when you switch focus you'll have sensitivities changing, which
will undoubtedly be confusing.

Would View as List/View as Icons apply per-pane?  If so then it will be
flipping values around constantly as focus changes between the pane
children.

Another big issue is how do you deal with focus?  Right now if you have
the side pane viewable and the focus is on it, then you hit ctrl-A to
select all, there is no doubt what you mean.  But if you have multiple
views open and your focus is not on one of those views and you hit
ctrl-A, what should happen?  You have the additional problem that it's
not always obvious which view even has the focus, because the focus
would be most easily determined by the user by seeing where the
selection is, and the selection might be outside the current viewport.
Or there may not even be a selection.  So there is a huge potential for
misunderstanding of the user about what will be affected by different
key bindings.

So, I think it seems like a cool idea.  But I don't understand how you
envision the UI to work, and I can't think of a good solution.  Until
really good ideas for UI are introduced, the coolness factor seems far
outweighed by all the problems that would be introduced into the user
interface.

/ Cody