W3C home > Mailing lists > Public > public-tracking@w3.org > November 2011

Re: "cross-site"

From: Jonathan Robert Mayer <jmayer@stanford.edu>
Date: Sat, 19 Nov 2011 13:03:24 -0800 (PST)
Message-Id: <5398B6AA-E2C8-4081-A707-E824EF1D302D@stanford.edu>
To: John Simpson <john@consumerwatchdog.org>
Cc: Ed Felten <ed@felten.com>, Mike Zaneis <mike@iab.net>, "<public-tracking@w3.org>" <public-tracking@w3.org>
*resending since copy+paste fail*

Suppose a first party doesn't want third-party resources on its site, so it sends web server logs to a third-party analytics service instead of embedding a script or pixel. So long as the data transferred is data the third party could have collected under the outsourcing exception, this text would allow the practice.

At a higher level, this text is just another instance of the principle (shared by many around the table, I believe) that we should not be (to the extent possible) drafting language that is technology-specific.

On Nov 18, 2011, at 1:19 PM, John Simpson <john@consumerwatchdog.org> wrote:

> Thanks, Jonathan.  Interesting proposal.  Can you please give me an example of what data a First Party site could transfer under the "may otherwise transfer data" language?
> 
> 
> On Nov 18, 2011, at 12:42 AM, Jonathan Mayer wrote:
> 
>> Agreed.  Between the discussion in Santa Clara, this thread, and these threads, I think we're very close to a consensus on first-party obligations.  Some time ago I drafted this text for the compliance document:
>> 
>>> First-Party Requirements:
>>> This standard imposes no requirements on first-party websites.  A first-party website MAY take steps to protect user privacy in responding to a Do Not Track request.
>> 
>> Here's what I would now propose:
>> 
>> First-Party Website Requirements
>> 
>> 1. Transfer of Data to a Third-Party Website
>> A first-party website MUST NOT transfer data to a third-party website that the third-party website could not collect itself under this standard.  A first-party website MAY otherwise transfer data to a third-party website.
>> 
>> 2. Additional Voluntary Measures
>> A first-party website MAY take additional steps to protect user privacy in responding to a Do Not Track request.
>> 
>> a. Example Voluntary Measures (Non-Normative)
>> […]
>> 
>> ...and then...
>> 
>> Third-Party Website Requirements
>> 
>> 1. Transfer of Data from a First-Party Website
>> If a third-party website receives data from a first-party website, the data is subject to the same collection, retention, and use limitations under this standard as if the third-party website had collected the data itself.
>> 
>> Jonathan
>> 
>> (tags: ISSUE-17, ISSUE-51)
>> 
>> On Nov 17, 2011, at 2:37 PM, Ed Felten wrote:
>> 
>>> It seems to me that there might be substantial agreement here.  As I
>>> understand John, he was positing two reasons for sending a DNT flag to
>>> first parties: (1) when DNT is enabled, first parties shouldn't
>>> circumvent the limits on third-party collection by collecting data and
>>> then sharing it with third parties, and (2) some first parties might
>>> choose voluntarily to go beyond what the standard requires when they
>>> see a DNT flag.
>>> 
>>> On Thu, Nov 17, 2011 at 3:28 PM, Mike Zaneis <mike@iab.net> wrote:
>>>> This is where there is a fundamental split amongst the parties. We had a
>>>> discussion several weeks ago about the first party obligations and I pointed
>>>> out that IAB and my member companies generally support the U.S. FTC position
>>>> that consumers don't expect first parties to be subject to such
>>>> restrictions.  Those positions have not changed.
>>>> 
>>>> Mike Zaneis
>>>> SVP & General Counsel, IAB
>>>> (202) 253-1466
>>>> On Nov 17, 2011, at 2:56 PM, "John Simpson" <john@consumerwatchdog.org>
>>>> wrote:
>>>> 
>>>> Shane,
>>>> I don't understand why we would say that a 1st party most likely will not be
>>>> subject to the DNT signal.  If we continue to use the 1st party/ 3rd party
>>>> distinction, it will likely (almost certainly) have different and probably
>>>> fewer obligations than a third party. It should still be subject to the
>>>> signal.
>>>> As a user I want the 1st party site to know that I have DNT configured.  As
>>>> a 1st party site operator I want to know a visitor has configured DNT and is
>>>> sending me the signal.  There will be some "musts", ie not sharing data from
>>>> a DNT configured user with 3rd parties, but if I am a responsible site
>>>> operator I may chose to go further in honoring the DNT request.  For
>>>> instance I might chose to not even include the visitor in my analytics. I
>>>> need to know if  DNT is configured and the way this happens is by being
>>>> subject to the DNT signal.
>>>> The obligations are different, but its important that we think of all sites
>>>> being subject to the DNT signal, once it is configured in the browser.
>>>> 
>>>> 73s,
>>>> John
>>>> On Nov 17, 2011, at 7:22 AM, Shane Wiley wrote:
>>>> 
>>>> Karl,
>>>> 
>>>> This statement is an attempt to remove the concern that a 1st party, which
>>>> will mostly likely not be subject to the DNT signal, does not have a
>>>> backdoor opportunity to pass user data directly to a 3rd party (aka -
>>>> closing a loop-hole).  3rd parties present on the 1st party's web site
>>>> should honor the DNT signal directly.
>>>> 
>>>> - Shane
>>>> 
>>>> -----Original Message-----
>>>> From: Karl Dubost [mailto:karld@opera.com]
>>>> Sent: Thursday, November 17, 2011 5:40 AM
>>>> To: Shane Wiley
>>>> Cc: John Simpson; Jules Polonetsky; Nicholas Doty; Roy T. Fielding; Mark
>>>> Nottingham; <public-tracking@w3.org>
>>>> Subject: Re: "cross-site"
>>>> 
>>>> 
>>>> Le 16 nov. 2011 à 23:30, Shane Wiley a écrit :
>>>> 
>>>> Alter statement to read "First parties must NOT share user specific data
>>>> with 3rd parties for those user who send the DNT signal and have not granted
>>>> a site-specific exception to the 1st party."  This will leave room for
>>>> sharing with Agents/Service Providers/Vendors to the 1st party -- as well as
>>>> sharing aggregate and anonymous data with "others" (general reporting, for
>>>> example).
>>>> 
>>>> I guess you mean
>>>> s/DNT signal/DNT:1 signal"
>>>> 
>>>> Trying to understand what you are saying.
>>>> 
>>>> 1. User sends DNT:1 to a website with domain name www.example.org
>>>> 2. www.example.org collects data about the user
>>>>   (IP address and categories of pages the user visits)
>>>> 3. Company Acme Hosting Inc. (a 3rd party) has access to these
>>>>   data NOT through the Web but through an access to the logs file.
>>>> 
>>>> 
>>>> What is happening?
>>>> 
>>>> 
>>>> --
>>>> Karl Dubost - http://dev.opera.com/

>>>> Developer Relations & Tools, Opera Software
>>>> 
>>>> 
>>>> 
>>>> ----------
>>>> John M. Simpson
>>>> Consumer Advocate
>>>> Consumer Watchdog
>>>> 1750 Ocean Park Blvd. ,Suite 200
>>>> Santa Monica, CA,90405
>>>> Tel: 310-392-7041
>>>> Cell: 310-292-1902
>>>> www.ConsumerWatchdog.org
>>>> john@consumerwatchdog.org
>>>> 
>>> 
>>> 
>> 
> 
> ----------
> John M. Simpson
> Consumer Advocate
> Consumer Watchdog
> 1750 Ocean Park Blvd. ,Suite 200
> Santa Monica, CA,90405
> Tel: 310-392-7041
> Cell: 310-292-1902
> www.ConsumerWatchdog.org
> john@consumerwatchdog.org
> 
Received on Saturday, 19 November 2011 21:04:04 UTC

This archive was generated by hypermail 2.3.1 : Friday, 21 June 2013 10:11:22 UTC