Re: Gnome-Shell - questions and opinions

[Adam Williamson <awilliam redhat com>]

On Sun, 2011-01-02 at 08:50 +1100, Bojan Smojver wrote:
> > For expose, I'm sorry that you don't like, but before we remove the
> > feature or make it optional, we need to understand why you feel it is
> > wrong.
> 
> The problem with expose is that it moves your windows around and resizes 
> them, many times unnecessarily.
> 
> When one uses workspaces, it is often not required to have windows that 
> overlap each other. In other words, one sets the windows on each worspace 
> so that they are fully visible. That's the point of having workspaces - you 
> get much more screen real estate. The eye and the brain "remember" the size 
> and the position of windows.

I'm somewhat worried by the seemingly exclusive focus of Shell design on
a workspace-based method of window organization. I don't use them, and
really don't find any advantage in using them. I set up my laptop with
workspaces based on the two monitors I use on my desktop, and really
tried to like it, just to try and drink the kool-aid for Shell; I didn't
feel that it gave me any discernible advantage at all, and I felt two or
three small disadvantages. I never once found myself 'naturally'
switching between workspaces, as I do find myself 'naturally' using new
UI elements I find really useful (like the sidebar in old Shell, or the
application run dialog in the overview of any Shell). I just carried on
switching between applications, and when one app happened to be on a
different workspace, I saw a 'workspace shift' animation.

No, I have no data, but I rather suspect a lot of users don't actually
open enough windows at once - and, particularly, enough *small* windows
which you can sensibly arrange in non-overlapping fashion on a typical
monitor - to benefit from workspaces, much less the ridiculous numbers
of workspaces lately being discussed on this list. A typical work
session of mine has only two windows - a terminal and my password
manager - which I could really usefully organize as an individual
workspace; obviously, a password manager and a console ain't a useful
work area. All my other windows are full-screen or close to it.

Are we really expecting everyone to voluntarily migrate to a new method
of working, whose benefits are probably small and likely obscure to
them? History suggests this is not likely to happen. Am I missing some
vital principle of a workspace-based system which would enable me to
take some advantage from it (what)? If so, I like to flatter myself that
I'm a vaguely savvy and informed user; do we expect others won't have
this problem? Or do we expect that my use case is sufficiently odd that
I'm not a useful test subject and most people really will have such a
set of windows as will be conducive to a workspace-based system? Are
there plans to somehow expose the greatness of workspaces to users to
mitigate it?
-- 
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Fedora Talk: adamwill AT fedoraproject DOT org
http://www.happyassassin.net