RE: ACTION-69: Renaming ISSUE-54

Vincent,

Correct - the order of superiority is "out-out cookie < DNT signal < explicit consent" (not sure this will always be typified as a cookie).

I believe a web-wide exception only model MAY ultimately become an option but I don't believe it's a valid starting point.

- Shane

From: Vincent Toubiana [mailto:v.toubiana@free.fr]
Sent: Tuesday, February 14, 2012 12:47 PM
To: JC Cannon
Cc: Shane Wiley; Jonathan Mayer; public-tracking@w3.org (public-tracking@w3.org)
Subject: Re: ACTION-69: Renaming ISSUE-54

I get your point, but I think a similar result can be achieved if the browser sets a web-wide exception based on the out-of band signal and then synchronize with the other devices.
On the other hand, I don't see how the problem I'm pointing (shared browser) could be addressed if out-the-band signal overrides DNT signal. Users would have to always remove all cookies.

I'd like also to recall that currently we have "DNT signal > Opt-Out cookie" (see action-52) and what you're proposing is "Opt-in cookie > DNT signal".
I think an approach based on web-wide exceptions could be more consistent.

Vincent

On 2/14/2012 5:08 PM, JC Cannon wrote:
+1

From: Shane Wiley [mailto:wileys@yahoo-inc.com]
Sent: Tuesday, February 14, 2012 8:04 AM
To: TOUBIANA, VINCENT (VINCENT); JC Cannon; Jonathan Mayer
Cc: public-tracking@w3.org<mailto:public-tracking@w3.org> (public-tracking@w3.org<mailto:public-tracking@w3.org>)
Subject: RE: ACTION-69: Renaming ISSUE-54

Vincent,

I strongly disagree that anything should trump an out-of-band explicit consent experience and I believe the group for the most part agrees on this stance.

- Shane

From: TOUBIANA, VINCENT (VINCENT) [mailto:Vincent.Toubiana@alcatel-lucent.com]
Sent: Tuesday, February 14, 2012 8:52 AM
To: Shane Wiley; JC Cannon; Jonathan Mayer
Cc: public-tracking@w3.org<mailto:public-tracking@w3.org> (public-tracking@w3.org<mailto:public-tracking@w3.org>)
Subject: RE: ACTION-69: Renaming ISSUE-54

Shane,

There are already tools to synchronize settings (bookmarks, tabs, history,...) on different browsers. I'm not saying that all browsers should implement site-specific exception this way, but that some could.
This is not a matter of granularity; it's just that (on some browsers) site specific exceptions should overrule out-of-band permissions (i.e. if no exception is granted out-of band exception should be ignored).


Vincent
________________________________
De : Shane Wiley [mailto:wileys@yahoo-inc.com]
Envoyé : mardi 14 février 2012 16:32
À : TOUBIANA, VINCENT (VINCENT); JC Cannon; Jonathan Mayer
Cc : public-tracking@w3.org<mailto:public-tracking@w3.org> (public-tracking@w3.org<mailto:public-tracking@w3.org>)
Objet : RE: ACTION-69: Renaming ISSUE-54

Vincent,

While it may be possible (read "MAY") to host more permissions through a single browser, in today's world where users move across devices at a fairly rapid pace this won't be taken up by publishers and users alike (until browsers solve online ID issues - see Mozilla's work here).  Until all online permissions are ubiquitously held on a web browser, out-of-band permissions will continue to be the primary approach to managing account details.  I don't believe the currently envisioned site-specific exceptions (or web wide exceptions) structure supports the granularity of permissions you're envisioning (nor should it - this is about DNT).

- Shane

From: TOUBIANA, VINCENT (VINCENT) [mailto:Vincent.Toubiana@alcatel-lucent.com]
Sent: Tuesday, February 14, 2012 8:09 AM
To: Shane Wiley; JC Cannon; Jonathan Mayer
Cc: public-tracking@w3.org<mailto:public-tracking@w3.org> (public-tracking@w3.org<mailto:public-tracking@w3.org>)
Subject: RE: ACTION-69: Renaming ISSUE-54

Shane,

I think the motivations listed in the "user managed site-specific exception" section are valid and I don't see why one could not implement a solution where all permissions would be hosted in the browser. Especially, users should be able to view quickly all the exceptions they have granted and that could actually avoid confusions.
Furthermore, users may not be aware of permissions that have been granted from the browser they're using (shared computer) and therefore should not be held responsible for all of them.

Vincent
________________________________
De : Shane Wiley [mailto:wileys@yahoo-inc.com<mailto:eys@yahoo-inc.com>]
Envoyé : mardi 14 février 2012 15:47
À : TOUBIANA, VINCENT (VINCENT); JC Cannon; Jonathan Mayer
Cc : public-tracking@w3.org<mailto:public-tracking@w3.org> (public-tracking@w3.org<mailto:public-tracking@w3.org>)
Objet : RE: ACTION-69: Renaming ISSUE-54

Vincent,

Specific to #2 - I don't believe attempting to host ALL online permissions via a web browser is an intelligent outcome.  By allowing race conditions to occur (consent vs. site-specific exception) we'll confuse everyone involved (users, site operators, regulators, etc.).  Nothing should trump an explicit consent experience.  If a user is not pleased with a previous permission they've granted, they should simply return to the source and revoke that permission.  This is a fair and appropriate expectation (users need to carry a modicum of responsibility in this exchange).

- Shane

From: TOUBIANA, VINCENT (VINCENT) [mailto:Vincent.Toubiana@alcatel-lucent.com]
Sent: Tuesday, February 14, 2012 7:09 AM
To: JC Cannon; Jonathan Mayer
Cc: public-tracking@w3.org<mailto:public-tracking@w3.org> (public-tracking@w3.org<mailto:public-tracking@w3.org>)
Subject: RE: ACTION-69: Renaming ISSUE-54

JC,

Web-wide exceptions have been considered precisely to address this issue.  These exceptions could be synchronized using tools provided by browser vendors.

In fact, devices synchronization should be addressed by browser vendors (there are tools to do that already) and we should not try to support a specific implementation as I see at least two possibilities:
1) A first implementation may use the FB account to synchronize the browsers settings (may require an API).
2) Some users will prefer to have their browser settings prevailing and ignore the FB account settings (or ignore all "out of band" exceptions). For instance, to prevent that a single FB account overwrites the settings of a browser used by several users.

Users may choose or configure their browser based on their priorities.

Vincent



________________________________
De : JC Cannon [mailto:jccannon@microsoft.com]
Envoyé : mardi 14 février 2012 02:04
À : Jonathan Mayer
Cc : public-tracking@w3.org<mailto:public-tracking@w3.org> (public-tracking@w3.org<mailto:public-tracking@w3.org>)
Objet : RE: ACTION-69: Renaming ISSUE-54

At this point I would like to hear opinions from other people.

If DNT breaks my relationship with companies I'll just leave it off. That would just make for a bad experience. Punting to browser vendors doesn't solve the multi-browser, user, computer issue.

JC
Twitter<http://twitter.com/jccannon7>

From: Jonathan Mayer [mailto:jmayer@stanford.edu]
Sent: Monday, February 13, 2012 5:00 PM
To: JC Cannon
Cc: public-tracking@w3.org<mailto:public-tracking@w3.org> (public-tracking@w3.org<mailto:public-tracking@w3.org>)
Subject: Re: ACTION-69: Renaming ISSUE-54

JC,

I share your view that requiring users to give explicit consent on each device is an issue.  We may want to make some accommodation in the exception API, or we may decide to punt to the browser vendors.

That's all, however, independent of the issue you raised: does Do Not Track impose limits on third parties when 1) the user has an account, and 2) the user is logged in?  I think the answer should be yes - the very same limits it imposes on any other third party.

Jonathan

On Feb 13, 2012, at 4:50 PM, JC Cannon wrote:

If I have an account with a company and I set my preferences a certain way, I don't want DNT overriding that, while I'm logged in. The alternative is to try to set an exception in every browser on every computer I use when I've already stated my position. To me that would be nightmare scenario.

No, a privacy statement in and of itself should not override DNT. However, I would like to see all privacy statements reference an organization's position on DNT, including when they feel that a member's preferences may override a DNT header.

JC
Twitter<http://twitter.com/jccannon7>

From: Jonathan Mayer [mailto:jmayer@stanford.edu]<mailto:%5bmailto:jmayer@stanford.edu%5d>
Sent: Monday, February 13, 2012 4:42 PM
To: JC Cannon
Cc: Shane Wiley; public-tracking@w3.org<mailto:public-tracking@w3.org> (public-tracking@w3.org<mailto:public-tracking@w3.org>)
Subject: Re: ACTION-69: Renaming ISSUE-54

Is it your position that a privacy policy link would be adequate to override DNT: 1?  If not, why can't we write that?

As for sites where the user already has an account, why grant them a special tracking privilege?  Any third party could put together a page for managing preferences.  (And several do.)

Jonathan

On Feb 13, 2012, at 4:38 PM, JC Cannon wrote:

I feel measuring the efficacy of a company's notice is out of scope here. I also don't feel we are focusing on social sites, but sites where the user has an and can manage their preferences in a centralized persistent fashion.

JC
Twitter<http://twitter.com/jccannon7>

From: Jonathan Mayer [mailto:jmayer@stanford.edu]<mailto:%5bmailto:jmayer@stanford.edu%5d>
Sent: Monday, February 13, 2012 4:33 PM
To: Shane Wiley
Cc: JC Cannon; public-tracking@w3.org<mailto:public-tracking@w3.org> (public-tracking@w3.org<mailto:public-tracking@w3.org>)
Subject: Re: ACTION-69: Renaming ISSUE-54

This is why standards of notice are important.  If "the user has already given Facebook consent" means Facebook asked explicitly for permission to identify the user on other sites and got an exception through the exception API, all's well.  But if Facebook merely links to its privacy policy at signup, no, that's not enough notice.  To the extent others believe it is, I'd like to hear why social networks deserve a special carveout from the ordinary rules.

Jonathan

On Feb 13, 2012, at 4:00 PM, Shane Wiley wrote:

Wouldn't the current Facebook structure be considered an "out of band consent" exception?  Meaning the user has already given FB consent to this data collection web wide?

- Shane

From: JC Cannon [mailto:jccannon@microsoft.com]<mailto:%5bmailto:jccannon@microsoft.com%5d>
Sent: Monday, February 13, 2012 4:55 PM
To: Jonathan Robert Mayer
Cc: public-tracking@w3.org<mailto:public-tracking@w3.org> (public-tracking@w3.org<mailto:public-tracking@w3.org>)
Subject: RE: ACTION-69: Renaming ISSUE-54

How about this:

1)    For social sites, manage social feature at that site or by logging out.
2)   For tracking and personalization opt-out use DNT.

JC
Twitter<http://twitter.com/jccannon7>

From: Jonathan Robert Mayer [mailto:jmayer@stanford.edu]<mailto:%5bmailto:jmayer@stanford.edu%5d>
Sent: Monday, February 13, 2012 3:19 PM
To: JC Cannon
Cc: public-tracking@w3.org<mailto:public-tracking@w3.org> (public-tracking@w3.org<mailto:public-tracking@w3.org>)
Subject: Re: ACTION-69: Renaming ISSUE-54

Without DNT, users would not be guaranteed the ability to opt out. That's the status quo.

On Feb 13, 2012, at 2:41 PM, JC Cannon <jccannon@microsoft.com<mailto:jccannon@microsoft.com>> wrote:
That looks like something that is handled by Facebook, which seems great to me. Why does DNT have to get involved?

JC

From: Jonathan Mayer [mailto:jmayer@stanford.edu]<mailto:%5bmailto:jmayer@stanford.edu%5d>
Sent: Monday, February 13, 2012 2:33 PM
To: public-tracking@w3.org<mailto:public-tracking@w3.org> (public-tracking@w3.org<mailto:public-tracking@w3.org>)
Subject: Re: ACTION-69: Renaming ISSUE-54

Social widgets would not "stop working."  Users certainly would not need to "re-enable [widgets] on every site."  To clarify, here's our Facebook mock-up.


Embedded widget:


<image001.png>



Consent option:

<image002.png>



On Feb 13, 2012, at 2:07 PM, JC Cannon wrote:

If we start making changes that degrades consumers' online experience or makes them work too hard to get their experience back to normal they won't use. As a consumer I want to just click  a button and be protected without everything else changing. If I enable DNT:1 and all of a sudden all my social widgets stop working or you tell me to "just" re-enable them on every site, I'm just going to turn it off.

Rob provided a good example of a site<http://www.heise.de/newsticker/meldung/Apple-will-deutschen-Patentstreit-mit-Motorola-in-den-USA-gewinnen-1433070.html> that manages social widgets (widgets at bottom of article). I like the way it works, but I wouldn't want to manage the settings on every site.

Responses to your suggestions below.

JC
Twitter<http://twitter.com/jccannon7>

From: Jonathan Mayer [mailto:jmayer@stanford.edu]<mailto:%5bmailto:jmayer@stanford.edu%5d>
Sent: Monday, February 13, 2012 1:12 PM
To: Geoff Gieron - AdTruth
Cc: JC Cannon; Jeffrey Chester; John Simpson; Justin Brookman; public-tracking@w3.org<mailto:public-tracking@w3.org>
Subject: Re: ACTION-69: Renaming ISSUE-54

I've seen a few suggestions that I'd disagree with on this thread.  Since they seem fairly independent, I'll break them out individually.

1) We should allow social widget personalization because some users might want it.

The same argument applies to any prohibited practice.  That's why the TPE specification provides an explicit exception mechanism.  If a user wants to see more personalized content from a third party or using third-party data, whether stuff their friends liked or (tracking-based) interest-targeted advertising, they can provide an exception.  We mocked up an exception flow for Facebook's social widgets at http://donottrack.us/cookbook/.
[JC] I feel the logged in state is different from standard third-party interaction. As a consumer I don't want to provide an exception our use the cookbook strategy for something I already control. If I don't want the functionality I will log out.

2) Social widget providers are first parties if the user has logged in since the user has a business relationship with the website.

I thought we had agreement in Santa Clara and on the list that third-party widgets become first party by means of user interaction, not logged-in status.
[JC] I agree the widget does not get first-party status, but they should still be able to personalize my experience, which I control. I don't see that as first party functionality.

3) Social personalization isn't the same as third-party data collection.

I don't follow this line at all.  Embedded social content is from a third party, and that third party collects data.  Moreover, the data collection practices currently used for social personalization rely not only on a unique ID-they rely on an ID tied to the user's identity.
[JC] Third parties have the ID information whether DNT is enabled or not. The header tells them their obligations to me as a consumer and whether or not they can process my data.

On Feb 13, 2012, at 10:40 AM, Geoff Gieron - AdTruth wrote:

JC - you raise a valid point on tracking vs social sharing.  One thing I would like to point out about this though is how logged in entities like Facebook already supersede existing browser solutions, such as Private Browsing (notice how Facebook still knows it is you on sites outside their network when you are suppose to be incognito to all entities).

So based on your ascertainment - I do agree that DNT should not disrupt the value of social sharing when you are logged in and remained logged AND Private Browsing/Incognito Mode should be what cuts even this level of detection by these types of services regardless of login. However - we are not standardizing Private Browsing/Incognito Mode so I'm concerned that the ability by a few customer facing companies like Google, Yahoo, Facebook, Amazon, AOL and Microsoft essentially may have the ability to completely work around all existing browser based solution offered to a consumer.

Is my concern valid to you?

Geoff Gieron
Business Development Strategist

<B6D349F0-DB69-481C-A4A8-5CC1CDE1C45E[87].png>

O:   +1.480.776.5525
M:  +1.602.418.8094
ggieron@adtruth.com<mailto:ggieron@adtruth.com>
www.adtruth.com<http://www.adtruth.com>


From: JC Cannon <jccannon@microsoft.com<mailto:jccannon@microsoft.com>>
Date: Mon, 13 Feb 2012 17:29:21 +0000
To: Jeffrey Chester <jeff@democraticmedia.org<mailto:jeff@democraticmedia.org>>
Cc: John Simpson <john@consumerwatchdog.org<mailto:john@consumerwatchdog.org>>, Jonathan Robert Mayer <jmayer@stanford.edu<mailto:jmayer@stanford.edu>>, Justin Brookman <justin@cdt.org<mailto:justin@cdt.org>>, "public-tracking@w3.org<mailto:public-tracking@w3.org>" <public-tracking@w3.org<mailto:public-tracking@w3.org>>
Subject: RE: ACTION-69: Renaming ISSUE-54
Resent-From: <public-tracking@w3.org<mailto:public-tracking@w3.org>>
Resent-Date: Mon, 13 Feb 2012 17:30:10 +0000

I'm not stating that data can be collected on me. I'm only stating that during a logged-in state a social site may personalize my experience based on my settings on the social site. The social site does not have the right to capture my browsing habits or process any data on me unless I interact with its widget.

JC
Twitter<http://twitter.com/jccannon7>

From: Jeffrey Chester [mailto:jeff@democraticmedia.org]
Sent: Monday, February 13, 2012 9:16 AM
To: JC Cannon
Cc: John Simpson; Jonathan Robert Mayer; Justin Brookman; public-tracking@w3.org<mailto:public-tracking@w3.org>
Subject: Re: ACTION-69: Renaming ISSUE-54

Let's have a further discussion on this.  Can you say what data can be collected from a user when DNT:1 is on as they access social services?



Jeffrey Chester
Center for Digital Democracy
1621 Connecticut Ave, NW, Suite 550
Washington, DC 20009
www.democraticmedia.org<http://www.democraticmedia.org/>
www.digitalads.org<http://www.digitalads.org/>
202-986-2220

On Feb 13, 2012, at 12:09 PM, JC Cannon wrote:

Jeff,

I disagree with your position, because I would still want that feature if I'm logged in and have DNT:1 enabled. This is part of the concern I have about making decisions for consumers that they may not want. If I can disable the article annotation by logging off from the social site, why bundle it with DNT taking away my flexibility?

JC
Twitter<http://twitter.com/jccannon7>

From: Jeffrey Chester [mailto:jeff@democraticmedia.org]<mailto:%5bmailto:jeff@democraticmedia.org%5d>
Sent: Monday, February 13, 2012 6:46 AM
To: JC Cannon
Cc: John Simpson; Jonathan Robert Mayer; Justin Brookman; public-tracking@w3.org<mailto:public-tracking@w3.org>
Subject: Re: ACTION-69: Renaming ISSUE-54

JC:

DNT:1 serves as a form of granular privacy protection.  If one has DNT:1 on, they don't want tracking process working--even if it means they can't find out their friends enjoyed reading your latest book!  Happy to discuss.

I would like to drill into this a little further. How would this apply to a logged in state? If I'm logged into a social site and reading an article I would be interested to know if people I trust from that social site enjoyed the article or not without necessarily letting people know that I viewed the article, unless I select the share button. I don't want to have to enable tracking just to see if my friends liked the article.

JC
Twitter<http://twitter.com/jccannon7>

From: John Simpson [mailto:john@consumerwatchdog.org]<mailto:%5bmailto:john@consumerwatchdog.org%5d>
Sent: Wednesday, February 08, 2012 11:53 AM
To: Jeffrey Chester
Cc: Jonathan Robert Mayer; Justin Brookman; public-tracking@w3.org<mailto:public-tracking@w3.org>
Subject: Re: ACTION-69: Renaming ISSUE-54

I agree that when a site acts as a third party it MUST not engage in targeting based on data gathered when it was a 1st party if DNT is enabled.


On Feb 8, 2012, at 8:43 AM, Jeffrey Chester wrote:



I don't think if DNT is enabled a third party should be able to engage in profile-based targeting that they have collected as first party, as Justin perhaps as proposed.  That would weaken user intent on DNT.


Jeffrey Chester
Center for Digital Democracy
1621 Connecticut Ave, NW, Suite 550
Washington, DC 20009
www.democraticmedia.org<http://www.democraticmedia.org/>
www.digitalads.org<http://www.digitalads.org/>
202-986-2220

On Feb 8, 2012, at 11:34 AM, Jonathan Robert Mayer wrote:



In the interest of clarity, I recommend we make two ISSUEs from ISSUE-54.

1) What can a first party do on its own website with provided information? I completely agree with Shane that this falls into the current first party proposal, and I expect we'll get consensus and close the ISSUE quickly.

2) What can a first party do with submitted information when it's a third party? We've already heard a range of views on this; I expect lengthy discussion and perspectives from many stakeholders before we close the ISSUE.

On Feb 8, 2012, at 7:20 AM, Justin Brookman <justin@cdt.org<mailto:justin@cdt.org>> wrote:
I think Sean's restatement of the issue is a bit ambiguous.  The key question is not whether a first party can alter its own websites and advertising on those sites based on data it collected as a first party.  It's about whether they can then leverage that data when they're in a third-party environment.

I was tasked with writing up language on this in Brussels, but upon reflection, my vision is already allowed for in the text:  a third-party may customize content or advertising on other sites based on data it had collected as a first-party.  Thus, Yahoo! can serve ads on the New York Times based on what I had done on the Yahoo! site (or registration information I had provided to Yahoo!) and Facebook can tell me what my friends like in a social widget when I go to the WashingtonPost.com<http://WashingtonPost.com/> --- as long as neither collects the fact that I went to NYT or WaPo (apart from exceptions like ad reporting, fraud, analytics) and certainly does not add that information to a profile about me.  The language in the draft currently allows for this.  However, I will try to put together some non-normative language on this today to make it clear.  I have heard the argument that this unduly favors first-party sites who have a lot of user data, but I also think the privacy implications are dramatically reduced when ads are influenced based on data that a party already has about you.

Shane, you had seemed to disagree with this idea in Brussels, so if you want to put forward a countersuggestion that's fine.  Alternatively, Tom had disagreed on one of the calls that Facebook should be allowed to personalize content based on data it had collected as a first-party, so he may want to proffer another suggestion.  I could see a stronger argument against allowing Yahoo! to use passively-collected data about what I read on the Yahoo! site rather than using affirmatively provided info, but I personally wouldn't draw the line there.  It's also possible this issue is currently being discussed elsewhere on the mailing list, but I have not remotely been able to keep up.



Justin Brookman

Director, Consumer Privacy

Center for Democracy & Technology

1634 I Street NW, Suite 1100

Washington, DC 20006

tel 202.407.8812

fax 202.637.0969

justin@cdt.org<mailto:justin@cdt.org>

http://www.cdt.org<http://www.cdt.org/>

@CenDemTech

@JustinBrookman

On 2/6/2012 10:10 AM, Shane Wiley wrote:
And the proposed answer, "YES", as this appears to capture the 1st party exception cleanly and we have other statements that disallow a 1st party from sharing information with 3rd parties when DNT:1.

- Shane

From: Sean Harvey [mailto:sharvey@google.com]
Sent: Sunday, February 05, 2012 5:27 PM
To: public-tracking@w3.org<mailto:public-tracking@w3.org> Group WG
Subject: ACTION-69: Renaming ISSUE-54

Hi all, apologies for the delay in submitting my action item.

ISSUE-54 is intended to get at the question of whether or not a first party is allowed to leverage their own data, including registration data provided by the user at a previous time, in the context of a DNT header being ON.

Keep in mind I am not intending to provide an answer, only to more appropriately rename the topic.

In light of this I propose the Issue be renamed:

"Can first parties customize their own websites or advertising based on their own user data when a DNT header is ON?"



----------
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<http://www.ConsumerWatchdog.org/>
john@consumerwatchdog.org<mailto:john@consumerwatchdog.org>



The information contained in this e-mail is confidential and/or proprietary of AdTruth. The information transmitted herewith is intended only for use by the individual or entity to which it is addressed. If you are not the intended recipient, you should not copy, distribute, disclose or use the information it contains, please e-mail the sender immediately and delete this message from your system.

Received on Tuesday, 14 February 2012 20:45:18 UTC