24738 2014-02-20 03:00:21 +0000 D3E spec should define if keypress is fired when preceding beforeinput is consumed by preventDefault() 2014-04-07 23:17:21 +0000 1 1 1 Unclassified WebAppsWG HISTORICAL - DOM3 Events unspecified PC Windows NT RESOLVED FIXED P2 normal --- 1 masayuki travil bzbarsky garykac mike www-dom public-webapps-bugzilla oldest_to_newest 100941 0 masayuki 2014-02-20 03:00:21 +0000 D3E spec should define if keypress is fired when preceding beforeinput is consumed by preventDefault(). I think that keypress shouldn't be fired because: * it's not necessary for the web apps which handle beforeinput * not dispatching keypress doesn't waste the resource (CPU, battery) 103510 1 garykac 2014-04-07 22:38:34 +0000 Since keypress is deprecated, there is no need to define how it operates WRT beforeinput. And defining how this deprecated event should interact with the new events may make it more difficult for browsers to implement the new behavior. So, it seems best to leave this unspecified. 103511 2 bzbarsky 2014-04-07 22:41:59 +0000 Given the amount of keypress usage out there, in practice it _will_ be used in combination with the new events, and we will need UAs to be interoperable in that setting. So we do in fact need to spec this. 103514 3 garykac 2014-04-07 23:17:21 +0000 I've updated the spec (ED) to mention that keypress (if supported) should be suppressed if keydown is cancelled. However, we are not attempting to specify all possible interactions between the legacy events and the new events - partially because we don't want to needlessly constrain user agents implementing the new events and partially because many legacy events are not consistent across all user agents. But you're right that it makes sense to include that information when it is straightforward - as it is with this bug.