Re: Down-scoping / narrowing of ISSUE-187: Right approach to exceptions

I'm totally fine with separating the exception consent standard and exception model discussions into separate ISSUEs.  That seems like reasonable organization. 

If I'm not mistaken, we have a new status for an ISSUE that is resolved but with a deep dependency on the outcome of another ISSUE.  That seems to me the right status here: if we cannot establish a rigorous consent standard, then I (and based on discussions, several others) would not be comfortable with the new exception model.

Jonathan 


On Wednesday, April 10, 2013 at 12:19 AM, Matthias Schunter (Intel Corporation) wrote:

> Hi Jonathan,
> 
> 
> while composing the agenda for today, I have realised that your concern 
> of an insufficient user experience
> for agreeing to exception (the race to the bottom) is somewhat 
> independent of ISSUE-187:
> 
> - In the "old" approach to exceptions, we need some guidance to prevent 
> that some BROWSERS
> store exceptions without asking the user
> - In the "new" approach to exceptions, we need some guidance to prevent 
> that some SITES
> store exceptions without asking the user
> 
> I see this as a valid concern for both approaches to exceptions (and 
> similarly also for the initial gathering of preferences).
> 
> As a consequence, I suggest the following way forward:
> - We create a new ISSUE on "Rules for sufficient user interaction to 
> ensure that a user actually agrees with an exception" (or a similar wording)
> and include this as OPEN in the database and the current TPE draft
> - I will try to close ISSUE 187 while stating that this new issue needs 
> resolving (but that we overall try to work
> towards the new approach to exceptions).
> 
> I hope that this addresses your concern of a "race to the bottom" while 
> still giving the group a stable path for the exception framework.
> 
> Would this work? If yes, I will send a mail to the announce list to 
> formally implement this down-scoping of ISSUE-187.
> 
> 
> Regards,
> matthias
> 
>