To Jason:
    Always on top is inefficient if it potentially interferes with 
another application's workspace. You also can't assume that the 
application always on top is on a side of the screen. Also, having the 
system automatically go to full screen when there are no windows in the 
task bar is also naive. Imagine the scenario, "GAIM goes full screen". 
Ugly ugly ugly...

    The visual habit that is developed to opt to the right side of the 
screen for a side-application is created by a comfort level we have with 
the mouse. I would assume that you are right-handed. As for people who 
are right handed, it is easiest to move the mouse away from you instead 
of towards you, opting you to initially start using the right side of 
the screen. When there is more than one side-app, the left side of the 
screen is generally used as opposed to lining up next to the other 
side-app. This is because we would like to focus on the center of the 
screen rather than the sides. Our eyes have a tendency to be attracted 
to light, having our center of focus decentralized would cause us to 
catch the space that is next to our monitors which would cause us to 
feel awkward when trying to give it our attention (our monitors do not 
emit light outside of their screens). The side-app we give the most 
significance too will probably take the right side of the screen.

    These techniques of organizing for users are common for 
multitasking. Several applications have been built to be side-apps, and 
the way we organize our desktop environment using them removes usability 
of some of its features. The most obvious one is the maximize feature. 
If you maximize your central task window, you will cover your side task 
window.
    There are also difficulties in optimizing space allocation of a side 
task. Namely, overshooting the screen and optimizing space-allocation 
(taking up the entire side of the screen).

    Certain features such as snapping have been developed in certain 
Windows applications (ex. Winamp, Trillian) to overcome this problem of 
ease of space allocation. You can also see it used internally in 
programs such as Adobe Photoshop. If applied to all applications, 
however, snapping could become frustrating. It has uses in 
side-applications docking to the side of the screen and smaller windows 
attaching to each other.

    Snapping to the border of the screen should make the space taken by 
the side-app ignored by maximizing another application. If someone's 
trying to get away from your addicting side-tasks temporarily, they 
might want to use the full screen. This problem is already answered by 
having multiple workspaces (thank god).

    It is my opinion that side-docking (this may include top/bottom) and 
maximizing within side-apps should be developed. The problem has been 
stated, now how would we go about solving it?

Brainstorm
-snap-to-edge while moving+resizing window (like trillian or winamp)
-individual dock-to-left, dock-to-right, etc. buttons on windows
-a docking function that is activated by some sort of mouse and/or 
keyboard activity, or some button, and while holding the left mouse 
button down, a side for docking is determined by the mouses location 
relative to a static point. (This idea came after the prior two, but 
before I started writing. In my opinion this outdoes the prior two).

So what I've discussed for this have been the desired functionality 
(specialization for side-applications), brainstorm of interactions to 
use desired functionality, and quite limited consequences of desired 
functionality.

The most important part of this message is that we recognize the 
side-application vs. maximize  functionality problem. The new question 
is, how do we solve this problem?

I bet you thought that was the only issue I was going to discuss. 
Actually, I'm going to discuss every issue that can e discussed.

Users usually end up clustering smaller windows outside of 
side-applications together. The problem is that each window is treated 
independently. Wouldn't it be nice if you could cluster a set of small 
windows and treat them as one window, or be able to minimize or view 
them all at the same time? With tabbed applications I'm running into 
this problem less often, but separate smaller windows are still good for 
comparison purposes.

What's most important about this is that we identify a tendency of 
group-dependency (users like clustering them) for smaller windows.
_______________________________________________
Usability mailing list
Usability@gnome.org
http://mail.gnome.org/mailman/listinfo/usability

Reply via email to