Though it's pretty weird to serial tiltX and tiltY for mouse input, we have to do this, since Windows send mouse platform messages for pen inputs and touch, which are latter wrapped into WidgetMouseEvent by Gecko. Thus, WidgetMouseEvent doesn't only represent mouse input but also pen and touch. Besides, it would be nice to match the hierarchies of both Widget****Event and their corresponding serializers.

Is this something we want to have in the next release/months, i.e. P2? Or is it closer to a nice-to-have (P3) speaking from now?

IMHO, if we are going to launch Pointer Events in any form (i.e., pre-release), then it's P3 (nice-to-have); However, if we're not launching Pointer Events, then this issue has nothing to do with the very next release ;P

In this stage, I think we can set this bug as nice-to-have

Thanks!

Comment on attachment 8814782 [details] [diff] [review] Part2: Serialize WidgetPointerHelper in IPC message > template<> >+struct ParamTraits<mozilla::WidgetPointerHelper> >+{ >+ typedef mozilla::WidgetPointerHelper paramType; >+ >+ static void Write(Message* aMsg, const paramType& aParam) >+ { >+ WriteParam(aMsg, aParam.pointerId); >+ WriteParam(aMsg, aParam.tiltX); >+ WriteParam(aMsg, aParam.tiltY); >+ // We don't serialize convertToPointer and retargetedByPointerCapture since >+ // they are temporarily variable and sould be reset to default. s/sould/should/

Updated the patch summary.

Updated the patch summary and comments.

Pushed by cbook@mozilla.com: https://hg.mozilla.org/integration/mozilla-inbound/rev/fd2c8004db8a Part1: Refine WidgetPointerHelper. r=smaug https://hg.mozilla.org/integration/mozilla-inbound/rev/eece9df96e36 Part2: Serialize WidgetPointerHelper in IPC messag. r=smaug

https://hg.mozilla.org/mozilla-central/rev/fd2c8004db8a https://hg.mozilla.org/mozilla-central/rev/eece9df96e36