Re: Additional security and privacy considerations?

On May 26, 2009, at 8:56 AM, Doug Turner wrote:

>
> On May 26, 2009, at 6:00 AM, Andrei Popescu wrote:
>
>>
>> I am sorry but I do not see above anything that answers my question.
>> You are asking to separate the goal from the realization of the goal.
>> To me this is entirely unrealistic in this case: the only mechanism
>> that has been proposed so far to realize the goal is some sort of
>> visual indicator. But whether this achieves the goal is debatable.
>> Also, what if the UA has a full screen mode (this was mentioned
>> before)? So what is an implementer supposed to do? For these reasons
>> I'm afraid that adding your goal to the spec will put our  
>> implementers
>> in an impossible situation. This is why I oppose adding your goal to
>> the spec.
>>
>> Thanks,
>> Andrei
>
>
>
> Thanks Andrei.  I think I agree with most of what you said.  As I  
> stated before, Mozilla will make up its own mind regarding UI in  
> Firefox and Fenenc.  We agreed that having an explict permission  
> dialog with the user before sharing geolocation, but I do not think  
> we would be down for having some flashing widget thing that tells  
> you that geolocation is happening.  My point of view is that we  
> shouldn't spec out stuff that is going to make most UAs non- 
> conforming, that a blinking LED that says "the browser is doing  
> something" hurts users and is dreadfully ugly, and "forgetting" the  
> user permission after some seemingly random time interval is a  
> really bad idea.

+1 This spec is about getting a location, not how it is implemented.

Thanks,
-- Greg