-- Truls Becken, on Mon, 23 Nov 2009 19:40:36 +0100: > Michael Weirauch wrote: > > > 2009/11/23 Humdinger <humdingerb@xxxxxxxxxxxxxx>: > > > >> I'd like to continue the discussion on resizing windows with > >> CTRL+ALT+click-on-border, analog to the existing CTRL+ALT+click- > >> anywhere for moving. > > > > As there seem to be concers about the clickability of the (thin) > > borders, just make the window resize on RMB-drag (additionally). > > The "overlay" would be an additional visual guidance on where to > > click, > > but isn't yet necessary making the resizing work. > > I strongly agree with Michael that ALT+LMB to move and ALT+RMB to > resize anywhere in the window is very user friendly. It is no > coincidence that most X11 window managers use this configuration by > default. At least "user friendly" and "X11" didn't met in a single sentence. Such a matter/anti-matter annihilation could rip freelists.org to pieces... (j/k) :) I'm not opposed to additionally having mofifier+RMB do resizing, but having the modifier+LMB-on-border as a fallback (or default, depending how you look at it) has some advantages: * for single mouse button users, obviously. * finer control (X, Y, X/Y resizing, or even moving the window up, left, up/left while resizing). I guess the modifier+RMB-thingy always resizes from the bottom-right corner (X/Y resizing) or you couldn't click _anywhere you like_ in a window for resizing (without introducing some hidden areas like shown with our overlay solution). * Q: If you modifier+RMB-click in the top-left corner and the bottom- right corner resizes, what do you do if your mouse hits the upper screen edge? Isn't it strange to have the mouse move at position X, but the only the bottom-right corner of the window moves along with it? > I think the overlay would become annoying pretty fast, though. I guess that has to be tested. Maybe fading it quickly to a barely visible overlay will help. Much depends on how visually pleasing and suble the effect could be made. But, as I said, "Does it make sense to explore this area further at this time, or will that have to wait until someone (who?, when?) finds out if there is an implemention that won't suck cpu cycles nor designwise?" Regards, Humdinger -- --=-=--=-=--=-=--=-=--=-=--=-=--=-=--=-=--=-=--=- Deutsche Haiku News @ http://www.haiku-gazette.de