Re: DeviceOrientation still broken on iOS8 Safari

On Mon, Jun 23, 2014 at 3:51 AM, Rob Manson <roBman@buildar.com> wrote:
> It seems that the DeviceOrientation API implementation on iOS8 Safari is
> still not compliant.
>
> Can anyone from Apple comment if this is likely to be fixed before the full
> release?
>
> Is this actually being worked on at all?
>
> Thanks.
>
> PS: I know this is not a mailing list for specific platform bug reports -
> but this is still a major issue holding back adoption of the
> DeviceOrientation API in the wild. And the release of this major iOS update
> seems a great time for the WG to ask politely for this to be addressed 8)

It is good to discuss this here IMO. I do not have access to the iOS 8
beta yet but I think you are suggesting that no changes exist from the
iOS 7 implementation?

Cross-device compatibility can tentatively be achieved across both
Android+iOS7 DeviceOrientation implementations with the techniques
described in [1]. For example, a demo built on that theory [2] works
very well across both Android and iOS browsers at this point without
having to resort to hacks. If that is not your experience please
report any issues at [3].

However, once you clear this first hurdle (i.e. general compatibility)
another issue still exists. iOS-based DeviceOrientation is not
'world-accurate' hence Apple's introduction of 'webkitCompassHeading'
and 'webkitCompassAccuracy' [4]. This makes developing world-accurate
3-axis-based DeviceOrientation web apps still too complex for web
developers across devices right now.

Echoing Rob's initial request, how can we help implementers further here?

br/ Rich

[1] http://dev.opera.com/articles/w3c-device-orientation-usage/

[2] http://richtr.github.io/threeVR/examples/vr_basic.html

[3] https://github.com/richtr/threeVR/issues

[4] http://lists.w3.org/Archives/Public/public-geolocation/2011Jul/0014.html

>
> roBman
>

Received on Monday, 23 June 2014 09:45:42 UTC