- From: Kostiainen, Anssi <anssi.kostiainen@intel.com>
- Date: Fri, 12 Apr 2013 20:01:46 +0000
- To: "justin.lebar@gmail.com" <justin.lebar@gmail.com>, Anne van Kesteren <annevk@annevk.nl>
- CC: "public-device-apis@w3.org" <public-device-apis@w3.org>
[ + Justin ] Hi Justin, Anne, On Apr 12, 2013, at 8:19 PM, Anne van Kesteren <annevk@annevk.nl> wrote: > On Fri, Apr 12, 2013 at 6:12 PM, Kostiainen, Anssi > <anssi.kostiainen@intel.com> wrote: >> On Apr 11, 2013, at 5:24 PM, Anne van Kesteren <annevk@annevk.nl> wrote: >>> Since there are so many opportunities for the UA to just say "fuck it" >>> having this method sometimes maybe throw seems more like a burden than >>> an actual help to developers. I recommend not throwing at all. >> >> This feature is shipping in Firefox for Android and Firefox OS. >> >> Have you received feedback from developers that this is a burden? > > No, but given that the user agent may terminate either way (including > without exception), it is better not to throw, especially as this it > is implementation-dependent. You only want to throw if it's traceable > to a clear developer error, here it's not at all clear why. Justin - Anne would like to see vibrate() not throw. Are you aware of any related issues with the current implementation https://bugzil.la/679966? > Also, I think the IDL should say "(unsigned long or sequence<unsigned > long>) pattern" The IDL change seems proper, and matches implementations IIUC. Justin? >>> Also, "pause"?! Is that intentional? >> >> This was fixed in the Editor's Draft, see "spin the event loop": >> >> http://dev.w3.org/2009/dap/vibration/ > > So why does that spin the event loop? To make the vibration pattern behave as expected. Or is there a more appropriate construct for this? > Having this method be synchronous seems really bad. The method is synchronous for the sake of simplicity, for the web developer that is. Thanks, -Anssi
Received on Friday, 12 April 2013 20:02:23 UTC