W3C home > Mailing lists > Public > public-tracking@w3.org > October 2012

Re: Multiple DNT Headers (ACTION-283, ISSUE-150)

From: David Singer <singer@apple.com>
Date: Sat, 13 Oct 2012 11:03:36 +0800
Cc: Thomas Roessler <tlr@w3.org>, Adrian Bateman <adrianba@microsoft.com>, Shane Wiley <wileys@yahoo-inc.com>, Jonathan Mayer <jmayer@stanford.edu>, public-tracking@w3.org
Message-id: <745C57C6-A714-408A-A811-6D4241A813E7@apple.com>
To: John Simpson <john@consumerwatchdog.org>

On Oct 13, 2012, at 5:32 , John Simpson <john@consumerwatchdog.org> wrote:

> On Oct 12, 2012, at 2:24 PM, Thomas Roessler wrote:
> 
>> On 2012-10-12, at 16:03 -0400, Adrian Bateman <adrianba@microsoft.com> wrote:
>> 
>>> This subject has nothing to do with Internet Explorer. The question is what happens if a broken user agent sends multiple DNT headers, which violates the definition of the DNT header (which MUST only appear once). Options include a) assume DNT:1; b) assume DNT:0; c) ignore all the headers; d) if the multiple headers all have the same value use that value otherwise one of the previous options; etc.
>>>  
>>> Personally I don’t think we’ll be able to enumerate and describe all the ways that people might find to not follow the technical requirements of the standard and so in general I think the spec should be silent on this type of situation.
>> 
>> +100
>> 
>>> If we do indeed decide to write something down for this then I agree with Shane’s proposal: If a server receives conflicting DNT headers, it MAY choose to treat the transaction as if no DNT header had been received.  The Server MAY choose to alert the user about possible user agent configuration issues causing multiple, conflicting DNT header signals to be received.
>> 
>> Sounds reasonable to me.
> 
> 
> I am puzzled.  If there is a conflict why wouldn't you treat as if DNT:1 were enabled?  If you went with treating as if no DNT signal, then at the least I'd think the server MUST choose to alert the user about possible configuration issues.
>> 
> 
> 

I am in agreement that we're hard pressed to produce a spec. which defines the conformant behavior to conformant signals, and I would suggest we leave defining recommended behavior to non-conformant signals until such time as we feel we have time on our hands.

David Singer
Multimedia and Software Standards, Apple Inc.
Received on Saturday, 13 October 2012 03:05:30 UTC

This archive was generated by hypermail 2.3.1 : Friday, 21 June 2013 10:11:36 UTC