This seems incorrect to me.
It does seem odd to me too. CCing the patch writer and reviewer.
sounds like a bug. we probably never QI only static cast to it.
Relevant code in the URL field. This is an easy change of simply adding nsIAccelerationListener to the list of supported interfaces.
Created attachment 547639 [details] [diff] [review] Add nsIDeviceMotionListener to ContentParent's NS_IMPL_ISUPPORTS Patch attached. nsIAccelerationListener was renamed to nsIDeviceMotionListener at some point after this bug was filed.
Comment on attachment 547639 [details] [diff] [review] Add nsIDeviceMotionListener to ContentParent's NS_IMPL_ISUPPORTS Nice detective work. Would you mind putting together a new version that follows these guidelines to make checking it in easy: https://developer.mozilla.org/en/Mercurial_FAQ#How_can_I_generate_a_patch_for_somebody_else_to_check-in_for_me.3f ?
Hmmm, I thought I'd already done that. What's my patch lacking from those directions?
Oh, weird, I could have sworn is was missing. I'll push this with the r=jdm annotation.