RE: skeleton Geolocation API - Error codes

Hi Andrei.

I think there was consensus that we should have timeouts as part of the
API as an error scenario with the ability to set the timeout period.

Was there something I missed that was a strong argument against?

Thanks.

Chris 

-----Original Message-----
From: Andrei Popescu [mailto:andreip@google.com] 
Sent: Friday, August 08, 2008 8:18 AM
To: public-geolocation@w3c.org
Cc: Shyam Habarakada; Chris Butler; Alec Berntson; Chris Prince; Aaron
Boodman; Doug Turner; Aza
Subject: Re: skeleton Geolocation API - Error codes

Hello,

I have updated the Geolocation API spec with the error codes we have
discussed earlier. Due to the fact that it wasn't very clear if
timeouts would need to be explicitly handled by this API, I have left
that error code out for now.

Anyway, since there haven't been any new major controversies lately, I
wanted to kindly ask you to have another look at the spec and let me
know if you think the API is defined well enough for its first version
(i.e. give it a "thumbs up" if you think it's fine).

Thanks,
Andrei

On Mon, Aug 4, 2008 at 7:20 PM, Doug Turner <doug.turner@gmail.com>
wrote:
>
> On Aug 4, 2008, at 11:07 AM, Shyam Habarakada wrote:
>
>>> -----Original Message-----
>>> From: Doug Turner [mailto:doug.turner@gmail.com]
>>>
>>> I disagree that we need  to or should expose http protocol specific
>>> error codes.  Otherwise sounds fine.
>>
>> Hi Doug,
>> Would you mind sharing with us your reasoning here? That would give
us
>> more context. Thanks
>
>
> I see no need to special case one class of geolocation providers by
exposing
> its specific errors codes.  We should hide these implementation
details
> behind general error codes, or by providing a separate optional
interface or
> set of attributes.
>
> Doug
>

Received on Monday, 11 August 2008 13:45:14 UTC