W3C home > Mailing lists > Public > public-device-apis@w3.org > February 2013

Re: [battery] Test suite updated

From: Mounir Lamouri <mounir@lamouri.fr>
Date: Fri, 22 Feb 2013 16:51:07 +0000
Message-ID: <5127A1FB.1020905@lamouri.fr>
To: Anssi Kostiainen <anssi.kostiainen@nokia.com>
CC: "public-device-apis public-device-apis@w3.org" <public-device-apis@w3.org>, Boris Zbarsky <bzbarsky@mit.edu>, Peter Van Der Beken <pvanderbeken@mozilla.com>
On 22/02/13 14:45, Anssi Kostiainen wrote:
> Hi Mounir, All,
> 
> On 7.2.2013, at 18.05, ext Mounir Lamouri wrote:
> 
>> I just pushed to editorial changes in the Battery Status API specification:
>> - the spec no longer marks the BatteryManager interface as
>> [NoInterfaceObject] given that there was no reason to do so and the
>> default behaviour is to not being marked as such;
>> - the spec no longer defines NavigatorBattery. Instead, it extends
>> Navigator with "partial interface Navigator".
> 
> I updated [1] the test suite [2] to match these changes to the spec.
> 
> battery-interface-idlharness.html [3] shows red on many tests now on Firefox. This seems not to be an issue specific to the implementation of this spec alone, rather it's more about idlharness.js' strict interpretation of the Web IDL spec. E.g. the DOM4 IDL tests [4] behave similarly on all major implementations I've tested.
> 
> Most of the failed tests are related to not throwing TypeError on getting properties on BatteryManager's prototype object. [NoInterfaceObject] was masking those issues earlier.

We have an ongoing work to implement WebIDL in Gecko. In the current
Firefox release (19), the Battery API isn't following WebIDL. However,
in Nightly, it should be working better. I only have 3 errors in Nightly
(instead of 13).

I CC'd two mozillians who might have more insights regarding why we fail
those tests.

Thanks for updating the test suite btw :)
--
Mounir
Received on Friday, 22 February 2013 16:51:34 UTC

This archive was generated by hypermail 2.3.1 : Monday, 23 October 2017 14:53:58 UTC