- From: Tab Atkins Jr. <jackalmage@gmail.com>
- Date: Wed, 14 Aug 2013 23:16:56 -0700
- To: "Robert O'Callahan" <robert@ocallahan.org>
- Cc: www-style <www-style@w3.org>
On Wed, Aug 14, 2013 at 11:04 PM, Robert O'Callahan <robert@ocallahan.org> wrote: > Microsoft introduced a large number of CSS properties controlling scrolling > and zooming in Windows 8: > http://msdn.microsoft.com/en-us/library/ie/hh920761%28v=vs.85%29.aspx > As far as I know, none of them other than touch-action have been proposed > for standardization so far. I've been planning to do a spec for most of these for *months*, but I've been waiting for Jacob Rossi to give the go-ahead from MS legal that they're clear patent-wise and whatever to standardize them publicly. > One of the features that we have use-cases for at Mozilla is the ability to > snap scrolling to land at specific offsets. One such use-case is touch-based > horizontal panning between pages of the home-screen in FirefoxOS. At the end > of the panning gesture the screen should show a single page, not parts of > two pages. An underlying physics model determines which page is selected > when the gesture ends, and the selected page snaps into place using > animation. Similar use-cases arise from many situations involving scrolling > through a list of items. > > The Microsoft CSS properties for this are -ms-scroll-snap-points-x/y and > -ms-scroll-snap-type: > http://msdn.microsoft.com/en-us/library/ie/hh920761%28v=vs.85%29.aspx#creating_content_snap_points > Their approach requires explicitly specifying a sequence of CSS lengths (or > percentages) representing allowable snap offsets. It seems to me instead the > snap offsets should be derived automatically from the layout of the > descendants of the scrolling container, so that changes to that layout don't > require manual updating of some property set on the container. > > So, here's my proposal: > https://wiki.mozilla.org/Gecko:CSSScrollSnapping > I made this up just now so I expect it has problems :-). The most obvious > issue right now is that it doesn't handle RTL well. Vertical writing may > also be a problem. Yes, this is one of my favorite features, and I agree with your analysis. ^_^ ~TJ
Received on Thursday, 15 August 2013 06:17:43 UTC