- From: Rigo Wenning <rigo@w3.org>
- Date: Tue, 06 Mar 2012 19:57:26 +0100
- To: public-tracking@w3.org
- Cc: Matthias Schunter <mts@zurich.ibm.com>
I think we mix up GUI challenges with protocol defaults. A GUI will always be able to offer a "*" exception, in javascript or chrome. IMHO we have to make sure that the browsers can distinguish the 3rd parties and react in a differentiated way. This is done by the described "per third party" way. And whether then the browser sends an exception to all third parties is the browser's (user's) choice. What Kevin raises is interesting. If a browser denies a permission for a certain third party, should he communicate that to the first party? This coincides with the discussion Shane is having about querying the DNT status. Concerning the GUI considerations, I don't think they should be made in the Specification, but I strongly encourage browser makers and others here to group together and pool their knowledge and perhaps write a Note (how-to) for such user interaction instead of everybody inventing their own. But this is just a suggestion and completely optional.. Rigo On Tuesday 06 March 2012 19:14:59 Matthias Schunter wrote: > I do not know whether, e.g., providing a "*" consent while then > dynamically declaring/displaying the third parties in use would be > called 'informed consent'. Such a solution would be a mix of "*" > exceptions plus TPL-like blacklisting of some third parties. > > Did I misunderstand something? > Does anybody have ideas how to resolve this?
Received on Tuesday, 6 March 2012 18:57:50 UTC