Re: editor's draft updated

On Thu, Dec 3, 2009 at 3:21 PM, Steve Block <steveblock@google.com> wrote:
>>>> Test 00133: call watchPosition() with wrong type for third argument. Exception expected.
>>> I don't think that any type passed for the positionOptions argument
>>> should cause an exception. The spec states that 'PositionOptions
>>> objects are regular ECMAScript objects ...' and all of its properties
>>> are optional. In JS, any type can be converted to a (perhaps empty)
>>> object. The implementation should simply check for the presence of the
>>> relevant properties, but never throw an exception. See the WebKit test
>>> linked above.
>>
>> Do you think that the specification should be more explicit about the above?
> That's probably a good idea.
>
>>>> Test 00031: pass if getCurrentPosition returns withing 100ms
>>>
>>> I'm not sure it's useful to test that the function returns within a
>>> particular time limit, as the spec doesn't specify such a limit. I
>>> think the important thing to check is that the callbacks are always
>>> invoked asynchronously, rather than from within the function call.
>>
>> The statement that this test is checking is "must immediately return",
>> though. Clearly the test isn't great but I can't think of another way to
>> check.
> Sure, but I think the intent of the spec is not to limit the time
> taken by the method to return, but to require that the location
> acquisition process and the callbacks should be asynchronous. Andrei,
> can you clarify?
>

Yes, the intention is to require the acquisition process to happen
asynchronously. The same wording exists in other specs, for example
here:

http://dev.w3.org/html5/webdatabase/#dom-database-transaction

Thanks,
Andrei

Received on Monday, 7 December 2009 15:34:52 UTC