- From: Tamir Israel <tisrael@cippic.ca>
- Date: Fri, 27 Jul 2012 12:46:26 -0400
- To: Chris Pedigo <CPedigo@online-publishers.org>
- CC: Shane Wiley <wileys@yahoo-inc.com>, David Singer <singer@apple.com>, Mike Zaneis <mike@iab.net>, Jeffrey Chester <jeff@democraticmedia.org>, "Roy T. Fielding" <fielding@gbiv.com>, Justin Brookman <jbrookman@cdt.org>, "public-tracking@w3.org" <public-tracking@w3.org>
Chris -- the ideal for me would not necessarily be to bind servers to respect non-valid signals, but rather to find a way to resolve signal conflicts. It's not apparent to me that the IE10 case is so clear cut (or will remain so in light of the proposed mechanism here). If, for example, IE wanted to retain their stance in light of the approach Shane and David proposed, they could quite simply implement a mechanism that prompted users to confirm their DNT-1 if it's challenged by a server in this way. At that stage, you have a clearly expressed DNT-1 user preference, and if a server wishes to continue ignoring that, I would argue they are opening themselves up to liability. Even without this clarification, I think it's ambiguous at best for a server to ignore a DNT-1 from an IE10 user after directly notifying the user their browser is sending this signal (James' point). So, instead of letting servers unilaterally deem a signal non-compliant, or setting off an escalation of arms, why not simply build in a mechanism for confirming user intention through the UA. But again, I'm happy to defer on this point until some other elements of the spec are resolved. Best, Tamir On 7/27/2012 12:32 PM, Chris Pedigo wrote: > Tamir, in response to your comment here: > > "The same cannot be said about a DNT-1 signal that looks 100% valid, but is being rejected based on the fact some ad networks have decided the UA-side settings are not up to par." > > I agree that parties shouldn't be able to make up reasons to reject the DNT:1 signal and still claim compliance. In this case, however, I think we're talking about servers rejecting a signal from a UA that is clearly not compliant with the standard. Nothing is made up here. It's very clear. In this case, I believe we should have flexibility to honor or not. Binding industry to acknowledge invalid UAs seems to be a dangerous path to go down. > > > > On Jul 27, 2012, at 10:48 AM, "Tamir Israel"<tisrael@cippic.ca> wrote: > >> The same cannot be said about a DNT-1 signal that looks 100% valid, but is being rejected based on the fact some ad networks have decided the UA-side settings are not up to par.
Received on Friday, 27 July 2012 16:47:26 UTC