Re: [Usability] Discoverable off-screen window dragging

[Willie Walker <William Walker Sun COM>]

When I was working on X Windows accessibility at Digital in the early 
1990's, I modified the popular window manager at the time (mwm?) to have 
an option to automatically put up scrollbars if the window was too large 
for the display.  It seemed to work well, though I cannot remember 
if/how the user was able to adjust the scrollbars using keyboard traversal.

The proposed mouse movement thing seems like a great idea that would 
work well for mouse users.  I've seen very similar behavior when an X 
Window server is too large for the physical display.  The behavior is 
not at all unexpected, but instead feels very natural.

On a related note, enhancing metacity's window movement behavior from 
the keyboard would be really nice.  If you press Alt+F7 to move a 
window, you can move it all over the screen using the arrow keys.  But, 
if you press Alt+Space and select Move that way, you cannot push the 
window vertically off the screen -- the window gets stuck on the title bar.

Will

Calum Benson wrote:
> Cc'ing gnome-accessibility-list too...
>
>
> On 28 Nov 2007, at 23:23, Hans Petter Jansson wrote:
>
> > I've been discussing accessibility/usability with visually impaired
> > users lately, and one thing that came up, and that I believe to be
> > low-hanging fruit, is the problem of windows being bigger than the
> > screen in one or both dimensions.
> >
> > This happens frequently for visually impaired users, since they
> > generally have very large fonts.
> >
> > I was told by one user that the way he worked around this was by going
> > to the control panel, choosing a smaller font temporarily, moving the
> > window, then setting the big font again. Of course, he was very  
> > happy to
> > hear about the alt+drag shortcut.
> >
> > Which made me wonder if there's a more discoverable way of moving
> > windows around when they're too big/partially off-screen.
> >
> > One idea that came up was automatically adding scrollbars to the
> > windows, but I don't see how that could work reliably, and it would
> > clutter the screen and be error-prone/hard to do technically.
> >
> > A better idea might be something like the following logic in the  
> > window
> > manager:
> >
> > IF window is focused AND
> >    pointer is pushing against the edge of the screen AND
> >    window has area off that edge of the screen AND
> >    user is not dragging
> > THEN
> >    move the window in the opposite direction of the edge being pushed
> >
> > So e.g. if you have a focused window which is partially off the
> > right-hand side of the screen, and you push your pointer against that
> > side, bumping into the edge, the window will move to the left until  
> > you
> > can see its right-edge frame. The rate of movement would be equal  
> > to the
> > number of pixels the pointer "wants" to move off-screen at each
> > increment. Only the focused window would be affected.
> >
> > I think this would be a lot more discoverable and useful for  
> > everyone -
> > not just visually impaired users - and it looks like all the required
> > information is available to the window manager, so it shouldn't be
> > terribly hard to implement.
> >
> > Thoughts?
> >
> > --
> > Hans Petter
> >
> > _______________________________________________
> > Usability mailing list
> > Usability gnome org
> > http://mail.gnome.org/mailman/listinfo/usability