Re: ACTION-86: Write a clarification of dnt:c to apply to never-tracked resources

On Feb 7, 2012, at 5:39 AM, Matthias Schunter wrote:
> I believe that the current header spec covers this case by means of
> the header response that says that a resource does not do any tracking
> (beyond what might be permitted for third parties under DNT;1).

That's not quite my interpretation. I believe the most recently revised response header proposal from Tom [1] uses Tk:1 to indicate compliance with the third-party requirements in the Compliance spec, but doesn't indicate that no identifiable data is retained at all (maybe it's retained for a security/fraud exception, for example). 

> This corresponds to your text except that your text "except in
> aggregated and anonymous form" is now replaced by a pointer to the
> compliance spec.

I agree that the text should ultimately be replaced by a pointer to the Compliance spec, but I think it should be to the outcome of ISSUE-119, rather than just to third-party compliance. Tk:n would have a distinct meaning from Tk:1, but we can come back to it when we discuss ISSUE-119 rather than right now if that's simpler.



> On 2/1/2012 4:55 PM, Nicholas Doty wrote:
>> Instead of dnt-cached:
>>> dnt-no-tracking = n
>> *dnt-no-tracking* indicates that this resource is not used for
>> tracking. The resource may or may not be cached but any information
>> gained by the server about the user's interaction will not be
>> retained, except in aggregated and anonymous form.
>> This text might instead explicitly refer to the definition of
>> "absolutely not tracking" (ISSUE-119) once we have text there.
>> Thanks,
>> Nick

Received on Wednesday, 8 February 2012 02:08:01 UTC