W3C home > Mailing lists > Public > public-device-apis@w3.org > July 2010

Re: JavaScript Permissions interface in WebApps

From: Doug Turner <dougt@dougt.org>
Date: Fri, 2 Jul 2010 09:46:14 -0700
Cc: W3C Device APIs and Policy WG <public-device-apis@w3.org>
Message-Id: <A20B9BC6-28D5-4F0F-A09C-30CAF0C677E9@dougt.org>
To: John Morris <jmorris@cdt.org>

On Jul 2, 2010, at 9:20 AM, John Morris wrote:

> Doug,
> 
> I appreciate that you have consistently taken the position that no privacy protections should be built into any API.


It isn't so much that "protection" shouldn't be built into an API.  It is that adding attributes to an API offers zero protection to the user by themselves and can be harmful by making assertions that can never be verified.


> It is unfortunate that you cannot be in London in a week for either the workshop or the F2F, so that you could express your views on that subject.  And yes, we did "talk to death" many of the issues in December 2008 at the Geolocation WG F2F.

I really do not want to miss the face to face, but i have another obligation.


> I make these points simply to assert that the fact that the Geolocation WG "talked to death" the idea of taking action to protect privacy (and rejected that idea) is not evidence that such action should be rejected today.  There are tough issues confronting DAP in this area, and I hope you and others can engage in the DAP discussions to articulate your perspectives on the issues.  I don't think it is enough to say to the DAP group "go read what we said in Geolocation."

It isn't probably enough and I never said that.  I said something like (or I meant), before you respond please go check out the current thinking on an issue very similar to this where the major browser venders mostly agreed.

If nothing else, looking at the the geolocation wg notes is a good starting point.


> It's not clear to me that privacy is or would be the underlying cause for browser vendors to fail to implement the DAP specs, but in any event I think that would be an unfortunate place to end up.  I hope that they will engage in the DAP process.

It isn't clear to me how much Firefox would implement of the DAP spec.  My current understanding is that we would implement very little. Much of the specifications are very specific to widgets, not web.  Some of the specs are not tied directly to the dom (like the early drafts of capture).  Some do not provide little value but high risk (like System Info).  The reasons, as you see, aren't because of privacy attributes you want to add, but more core problems.

I hope that helps!
Doug Turner
Received on Friday, 2 July 2010 16:46:49 UTC

This archive was generated by hypermail 2.3.1 : Monday, 23 October 2017 14:53:45 UTC