- From: Masayuki Nakano <masayuki@d-toybox.com>
- Date: Sat, 14 Sep 2013 21:04:51 +0900
- To: Rick Byers <rbyers@chromium.org>
- CC: www-dom@w3.org
On 2013/09/13 23:11, Rick Byers wrote: > Thanks for response. So in what scenarios do you envision a UA ever > setting deltaZ to non-zero? Is this just future-proofing for some > not-yet-popular hardware? On Mac OS X, native mouse scroll event has deltaZ property. However, as far as I know, it's always zero with normal mouse/touchpad. Additionally, there is no document about the deltaZ's sign on Apple's document. Therefore, browsers shouldn't set the value to DOM WheelEvent's deltaZ until somebody finds actual device which causes non-zero deltaZ value. > > FWIW, I'd LOVE to see D3E define a zoom event. If Indie UI > UIManipulationRequestEvent gets traction with implementations then that > would be even better. In the meantime we continue to get requests to > solve this scenario now ;-) > > Thanks, > Rick > > > On Thu, Sep 12, 2013 at 10:22 PM, Masayuki Nakano <masayuki@d-toybox.com > <mailto:masayuki@d-toybox.com>> wrote: > > I don't think browsers should set deltaZ even when the default > action is zoom-in/out. > > The spec just said that "The user's *environment* might be > configured to associate zooming with rotation along the z-axis.". > > So, zooming by Ctrl + y-axis wheel rotation is *just* a default > action of some browsers. So, web applications should be able to > handle the raw level action such as "Ctrl + y-axis wheel rotation" > rather than operation for zoom. > > If web application needs to handle zoom action, D3E or UI Events > should define "zoom" event. > > > On 2013/09/12 22:56, Rick Byers wrote: > > Hi, > We'd like to use the 'wheel' event in blink to communicate zoom > requests > from the user, distinct from pan (eg. on a trackpad 2-finger > scroll vs. > pinch). The spec > (http://www.w3.org/TR/DOM-__Level-3-Events/#events-__wheelevents > <http://www.w3.org/TR/DOM-Level-3-Events/#events-wheelevents>) has a > suggestion to use deltaZ for this: > > "The user's environment might be configured to associate vertical > scrolling with rotation along the y-axis, horizontal scrolling with > rotation along the x-axis, and zooming with rotation along the > z-axis." > > I was looking into doing this in blink > (http://crbug.com/289887), but it > seems no other browser behaves this way. IE10 and Firefox appear to > instead use deltaY with the ctrl modifier set for zooms (pinch on a > trackpad). Any advice on how this should be handled? With IE and > Firefox already in agreement, I think it's better to follow > their lead > then to follow the suggestion in the spec. Perhaps the suggestion > should be updated to match implementations? > > One problem with either approach is that it's not clear how an > app can > detect that a device that supports zooming distinct from panning is > being used. For example, a mapping site probably wants the > wheel on a > physical mouse to control zoom (since panning can be > accomplished with > click and drag), but panning on a trackpad to pan the map instead of > zoom it (since pinching can zoom it). Any tips on how a UA should > advertise that zooming is supported? It's unfortunate, for example, > that deltaZ isn't permitted to be undefined when not supported > by the > input device. > > Thanks, > Rick > > > > -- > Masayuki Nakano <masayuki@d-toybox.com <mailto:masayuki@d-toybox.com>> > Manager, Internationalization, Mozilla Japan. > > -- Masayuki Nakano <masayuki@d-toybox.com> Manager, Internationalization, Mozilla Japan.
Received on Saturday, 14 September 2013 12:05:19 UTC