← Back to team overview

unity-design team mailing list archive

Re: Thoughts on Unity design

 

On Mon, May 23, 2011 at 8:38 PM, David Regev <david.regev@xxxxxxxxx> wrote:
> I have yet to see an actual study arguing against the efficiency of global
> menus, not even for large modern monitors. If you want to take the return
> trip into account and other such factors, feel free to do so. The
> preliminary calculations show that global menus will still be vastly more
> efficient (even taking the return trip into account). You’re not going to
> get very fat picking nits here. You really need to do some studies. Record
> people using both types of systems and measure the average amount of time
> taken up by using each type of menu. I would love to see such a study. I
> might even be able to help. That’s how interested I am in seeing the
> results.
>
You forgot two things:
According to the base formula if you double D and W the time stays
constant, but as I said this can't be correct. There is a certain top
speed pixel/s or whatever and it increases with D no matter the W. The
formula ignores that.

Multi-tasking. From my experience this is the only issue you'll notice
in your day to day activity without a stop watch ;) However I also
noticed that it isn't that bad because: I hardly ever use menus at all
(which I find pretty ironic considering our discussion and brings us
to point two below).

I agree, we'd definitely needed such study if we wanted to make the
best decision. But the reason I'm against the global menubar isn't
just that I argue traditional estimates comparing Mac and Window menus
are no longer that accurate (which only means that the difference is
less, not that it's necessarily the reverse...) my two major arguments
are following:

Global menus are more difficult to understand and use for new and
inexperienced users.
This is ignoring the Unity hover menu (which also makes the menu
slower to access and invalidates Fitts's Law in Unity). I say this
because of my own testing with OS X. I've written more about this in
other threads here.

Secondly, the global menu is not flexible and may prevent app
developers from making even better use of the advantageous area at the
top of the screen. This can be tabs for browsers or toolbars for
editors for example. Such interfaces generally don't really exist
right now (there's Chrome, Firefox and Photoshop) but I'd love to have
them. I kind of hack my own "Fitts desktop" by removing window
decorations and putting windows to the screen edges. That way I can
even have applications side by side, all sharing the top screen edge.
A desktop with a forced and static top panel prevents me from doing
that.
When working with images I'd like to have all the tools and menus on
all four screen sides, the content in the center. Photoshop comes
pretty close to that, it has no title bar, but also no hidden menu and
no windows to speak. It's just content and the tools you need in the
places it makes the most sense. Here both the Unity panel and the
launcher would get in the way. Hence why I also think the "sidebar"
should be movable to the bottom but that's offtopic. Anyway, in
Windows PS is even more flexible than OS X because it can put anything
at the top, not just text based menus.

I'm all for making use of the screen edge but in my opinion until app
developers start doing that there isn't really much improvement
possible. People want to use their apps efficiently, they don't care
about the OS. Therefore occupying all four corners and quite a bit
more than half the screen edges for OS level elements like GNOME 2 did
is a terribly "selfish" endeavor. Unity is quite a bit better here but
I'd like to note that the top bar and the left bar are far more
prominent and important than right and bottom, mainly because we read
left to right, top to bottom... Making this forced global menu at
least we can guarantee that the top screen edge isn't wasted entirely
but what if the application doesn't need the menu? What if it only
needs a very short menu and leaves plenty of room next to it? What if
the tool or tabbar is used much more frequently than menu items? What
if the user wants to decide for himself which of the bars he'd like to
access faster? What about having two app windows side by side?
By reducing the static elements of the OS we'd give developers the
chance to come up with better solutions to these questions.



Follow ups

References