Re: Control Center Behavior

[Malcolm Tredinnick <malcolm commsecure com au>]

Let me put on my "I'm a power user, and I agree with both sides of parts
of this argument" hat. I expect this to in no way affect the final
outcome, since I am not in the target audience, but I would advocate
leaving the shell version of the control centre available as it is now
-- not more visible, not less.

If people want it more visible for their own use, they can put a
launcher on the panel.

On Thu, Jul 18, 2002 at 12:54:54PM -0400, Havoc Pennington wrote:
> The problem with the shell is that it just doesn't make sense to
> special-case the launching of particular kinds of thing.

It does to me -- see below. :-)

[...]
> The case where I'd be wrong is if preferences are special or different
> in some way. i.e. I guess you could argue that the "shell" is an
> application, and the control panels are dialogs of that
> application. So then we'd just drop the menu items for preferences.
> 
> But in short we need to identify:
> 
>  - is the problem with using nautilus/menus really a problem 
>    that applies to all apps; if so, it's broken to break 
>    out or special-case preferences
>  - is the problem something specific to preferences, if 
>    so the shell could make sense

Very personal opinion about why the shell makes sense to me: Preferences
are the one thing that Nautilus does that I use. For every other task
provided by Nautilus that I use, I find other ways preferable.

The control centre graphical interface is more useful to me than the
menus because it is faster to use. I was trying to work out why this was
the case yesterday and my office mate thought I had lost my marbles
because I was opening and closing preferences like a madman. I _think_
it comes down to two things:

	(i) the icons are a bigger target to aim for than the menu
	selection, so I can hit it faster, and

	(ii) on average, the new window that gets opened after I select,
	say, "Background" is closer to the control centre window (be it
	Nautilus or the shell version) than it is to the menu item (I
	don't have "pile all my windows into the top left corner" turned
	on in my window manager, since I want it to be usable). So
	again, it's slightly faster and feels easier to get to the "next
	step" in the process of configuring things.

Now, I guess it's significant that, as a rule, I do not use the mouse
very much at all. It is much faster for me to open applications and
navigate around my multiple desktops and windows using keyboard
shortcuts, since my hands are then always in the one place whereas my
mouse requires me to relocate my hands significantly and find something
that often moves just because I needed that space for a sheet of paper a
second ago.

Configuring my preferences is something that I need to use the mouse for
and I don't really mind that, since I do it so rarely. When I do get
into "configure" mode, however, I tend to change a few things at once,
so I prefer it to be fast. Therefore I have a launcher on my panel that
starts the control centre when I need it and away I go.

[...]
> (Let me also echo Seth's sentiment that we should in no way listen to
> polls on gnomedesktop.org. ;-)

Agreed. It was 21 votes total in this case, after all. Hardly a
significant statistical sample (and just getting more votes won't leave
them immune to other sample problems).

> We all need to realize that if we want to succeed on the desktop, we
> are going to be coding for a silent majority; and the only way to get
> their needs right is to follow usability heuristic rules, and go out
> and actively user test and solicit feedback.
> 
> Of course we should use hidden prefs and the occasional compromise and
> so on to keep developers comfortable in the environment. But IMO we
> are after users who are entirely silent in all online hacking forums.

I was with you right up until that last part, Havoc (and that is the
reason for this post). A disturbing trend has formed is that of viewing
what you call "developers" are to be accomodated via work arounds and
are not in the main target group. I would much rather they were
considered to be a primary audience as much as the other categories of
users. While more advanced users _can_ use work arounds in extreme
cases, why should we have to? After all, the current medium and advanced
users user GNOME _now_ -- not in some hypothetical future. Keeping them
on the radar during design discussions is surely not too hard to do.

Nevertheless, the 2.0.x releases are not the time to worry about this
since the targets have been set out, so I will stay within site of the
party line.

Malcolm