Ouch, I do recant. Yes, a gutter is fine, let's keep it simple. The fact that my approach allowed the scroll bar to be centered over the box should speak for itself, and specifying the width of the scrollbar as a negative number is, well, not quite that intuitive and useful either.
I may have too much Haskell in my veins at the moment. Sebastiaan On 29/01/2008, why the lucky stiff <[EMAIL PROTECTED]> wrote: > > On Tue, Jan 29, 2008 at 11:34:25AM +0100, Sebastiaan Besselsen wrote: > > > > flow do > > > title "Re: A very long story" > > > ... even more etc ... > > > scrollbar left => -10 do > > > ... whatever goes here only shows when the scrollbar is gone ... > > > end > > > end > > > > > > > The point is that this would remind you that the scrollbar is going to > blot > > out an entire area when it appears, and you know exactly what that area > is > > going to be. > > A `scrollbar` method that takes a block? You definitely have higher > aspirations > for Shoes than I do. I especially love the negative ten! It is, > shall we say: awexome? > > Folks, now now, Shoes says: > > I. Only vertical scrolling matters > II. Vertical scrollbars are expected to automatically appear > -- (a precedent set by browsers) > a. Unless :scroll => false. > b. When scrollbars come on, don't affect the layout. > -- (the jumpiness of html) > III. Avoid extra methods at all cost, since it requires extra > space in the user's short term memory. > <http://hackety.org/2007/08/15/oneLinersAreCrucial.html> > -- A. Whitehead: By relieving the brain of all unnecessary work, a > -- good notation sets it free to concentrate on more advanced > problems. > IV. Keep access to widgets and chrome at a minimum, since they > require more platform-specific code and lead to bloat. > V. Shoes is tiny. A tiny toolkit. > > Now I really think you're all starting to come around to what is > surely a fine, fine compromise. Leave the gutter be. > > _why >
