Hi I can relate to the worries and would like to emphasize that this has been done before in multiple UI frameworks. In Montage (montagejs,org) we implemented a pattern claimPointer/releasePointer/surrenderPointer inspired from Cocoa's responders (first responder / responder chain) but adapted to classic DOM events, it helps solve problems like a slider in a scroller, which is a good test case. I can completely see the numbers of line of code that would be reduced using the capture API but know from experience that there are a lot of subtle use cases that are possible now and we've heard the concerns. We all know code speaks, especially to developers, so I would suggests, and it might already be done, to build and code a list of use cases and implement it with the spec to make sure they work on not, this could solve a lot of back and forth and be part of the spec? My $0.02 Benoit On Apr 9, 2013, at 2:56, Scott González <scott.gonzalez@gmail.com> wrote: > On Tue, Apr 9, 2013 at 8:48 AM, Brandon Wallace <brandon.wallace@yahoo.com> wrote: >> Stopping bubbling should be rare. > > I agree. We have a rule in jQuery UI that none of our widgets are allowed to stop propagation. The only exceptions we make are for cases where we want to act as if an event never occurred, such as stopping a click event after a drag.
Received on Tuesday, 9 April 2013 19:17:37 UTC