Tracking Protection Working Group Teleconference

12 Feb 2014

See also: IRC log


hefferjr, Ninja, +1.858.229.aaaa, dsinger, JackHobaugh, WileyS, Fielding, eberkower, +1.323.253.aabb, kulick, Ari, justin, AWK, sidstamm, Wendy, MECallahan, schunter, Chris_Pedigo, SusanIsrael, Chapell, [FTC], LeeTien, [Apple], hober
moneill2, robvaneijk, CarlCargill, npdoty
justin, schunter, CarlCargill
kulick, ninja


<trackbot> Date: 12 February 2014

<ninja> aaaa is WileyS

Confirmation of scribe and caller-identification

<eberkower> I wear a

<eberkower> brace on my wrist

<eberkower> sorry

<kulick> i'll scribe

<ninja> scribenick: kulick

<justin> scribenick: kulick

ISSUE-151: UA requirement to handle exceptions

justin: provide update on CFO and then go to 241 and 240
... and then talk about process moving forward
... we will take a break at closing out TPE and provide time to clean up draft and call out any issues with it

<ninja> link to results page: https://www.w3.org/2002/09/wbs/49311/tpwg-exception-151/results

justin: update on CFOs
... 151
... one of the hardest... strongest opins both ways
... option A seemed to have least strong obj
... option B we thought had strong arguments.... option c strong args as well mainly from dsinger on no tech need
... issue 153

ISSUE-153: Limitations for add-ons

<ninja> link to results page: https://www.w3.org/2002/09/wbs/49311/tpwg-addons-153/results

justin: least strong to adding sentence from shane and david singer to make UAs share responsibility

<dsinger> issue-241?

<trackbot> issue-241 -- Distinguish elements for site-internal use and elements that can be re-used by others (1/3) -- open

<trackbot> http://www.w3.org/2011/tracking-protection/track/issues/241

ISSUE-241: Distinguish elements for site-internal use and elements that can be re-used by others (1/3)

justin: two issues that will be going to CFO
... 241

<ninja> wiki link: https://www.w3.org/wiki/Privacy/TPWG/Proposals_on_elements_for_1and3_party_use

justin: 3 topins currently, but i thought it was narrowed down to two
... usage as 1st and 3rd party to be provided OR no change from Roy
... i dont think the proponents of opt 2 were willing to withdraw

<ninja> yes, Nick's proposal is no longer valid

justin: we will announce the CFO on this one tonight

shane: 2 weeks ago i mentioned 143

<dsinger> issue-143?

<trackbot> issue-143 -- Activating a Tracking Preference must require explicit, informed consent from a user -- closed

<trackbot> http://www.w3.org/2011/tracking-protection/track/issues/143

shane: i am expecting to turn that in today
... how does that work with the existing timeline

justin: we discussed on call last week
... however i know you werent there
... it was decided to be closed against TPE and moved to compliance
... thot could be explored in TPE 2.0

shane: is that a formal finding?
... can you put that in writing so I can object

justin: i'll do that
... if you have text or idea, please bring before the group

shane: ok... i'll do that... its unfortunate to deal after the fact

justin: sporry about that... i was going to remind, but forgot... there was some discussion on SNV calls and we looked at the minutes...
... we read as a decision not to pursue in the TPE at that time.
... i am personally interested and look forward to your idea

ISSUE-240: Do we need to define context?

justin: finally, discussion on context
... isuee 240

<ninja> wiki link: http://www.w3.org/wiki/Privacy/TPWG/Proposals_on_the_definition_of_context

justin: reach out to try to merrge some proposals
... 1, lang from Roy has been amended
... lang from Rob ....
... an chris PEdigo

and then one from Chris mejia

<fielding> I have been using it normatively since day 1

dsinger: when we defined tracking we decided that we wouldnt normatively used elsewhere but only to explain to the user what the protocol roughly does
... but not a formal def in other parts of hte doc

justin: i dont recall no normative consequences

<fielding> specifically, http://www.w3.org/2011/tracking-protection/drafts/tracking-dnt.html#expression-format

<dsinger> Matthias said several times it could be (would be?) informal, just for the purposes of explaining to the user and introducing the TPE

justin: ration for defining tracking is to provide what tracking means in normative terms (not sure I got this right)

<fielding> http://www.w3.org/2011/tracking-protection/drafts/tracking-dnt.html#TSV-1

matthias: 2 places int he cur spec... is where you mention DNT:1 and second is where site specifies no tracking

justin: that is our understanding of the co

<dsinger> OK, so we're somewhere in between

justin: the term does have imp meaning

<fielding> (hmm, looks like I have an old anchor name on that section)

dsinger: (didnt capture)

<fielding> again, totally not interested in discussing issue-5 again

schunter: user can express and and how server responds is diff

<dsinger> we're in between in the sense that multiple compliance regimes will define differently what turning off tracking means, and hence what tracking is

justin: we are not re-opeing issue 5, just discussing understanding

<dsinger> thanks, I think I am about as clear as an in-between state can be...

justin: user mentions what they want and server responds and the compliance handles what that means
... we have 4 options
... going to CFO today

<fielding> different compliance regimes might permit tracking in the presence of DNT:1, but it is still tracking in the eyes of the user.

justin: 2 weeks to re4spond
... primary goal: get TPE out

Announcement of plan for the next 3 weeks

justin: then turn back to compliance per Oct poll
... chairs will decide how soon to take up compliance
... decided for longer break due to IAPP and de-compress for folks...
... after break will take up open issues for compliance doc
... not sure of issues ordering. if you have thots, pls share
... questions?

pedigo: i am fuzzy
... on the next steps for the tpe
... will there be public comment, testing in real world?

justin: here is newbie chair understanding...

<fielding> the REC process is an infinite loop

justin: look over doc, identify errors

<wseltzer> W3C Process, Last Call

justin: then, group decides to put for public comment
... this is high profile event
... then make decisions about any responses
... then go to last call.
... wendy, sound right?

<Chapell> Why not wait until we've actually obtained learnings from last call before moving forward with compliance?

wendy: putting to pub comment is the last stage
... group agrees if ready for public review (i.e. last call)
... then process comments from public
... adress each of the public issues
... then call for imp review
... do we have some we can test... does it work as intended

<WileyS> +1 Alan - could we take a more meaningful break between TPE completion and C&S start? This will give us a bit more time get testing going on the TPE to inform C&S discussions.

wendy: then move to formal recommendations
... then we go back to charter
... charter expects us to produce TPE AND compliance
... further stages we will push compliance doc to same level

pedigo: since hybrid process, which TPE will go to last call versus those that wont

<Chapell> Charter does not proscribe timing --- in other words, we don't need to start compliance prior to obtaining learnings from getting TPE to last call

wendy: this is why i posted some
... of this
... (scribe not getting all this)
... there are stages after last call where it goes back to group and director before being finalized
... and we expect to have both tpe and compliance

pedigo: i'm still fuzzy
... i expect we will get a lot of comments on compliance
... does public comments get addressed by group

wendy: yes

pedigo: calls for imp, do you feel TPE is enough or you need compliance ?

schunter: we need to show inter-operability

pedigo: so we need to wait?

schunter: yes, i feel that

chapell: thanks justin

<WileyS> I think its the other way around - speaking as someone who will actually implement

<ninja> We don't want to rush Compliance. But steadily taking up the process again.

chapell: want to understand process to rushing towards compliance doc... concern of closing compliance issues without resolving comments to TPE

<WileyS> We should take some time (a few months?) to get testing up and running on TPE and allow those realities to inform the C&S discussion.

chapell: this is a practical concern of making decision with incomplete info

justin: the group will have some info that the group needs to consider
... chairs should be open to new info

<WileyS> Justin - to be fair we've not implemented so we haven't had new info to inject back into the discussion

justin: mission of w3c, should be standardize way to comply

chapell: is this w3c decision or working group decision to move forward

justin: Oct poll decision is what we are implementing now

chapell: charter doesnt req immediately do this

<ninja> Chairs' plan on moving forward: http://lists.w3.org/Archives/Public/public-tracking/2013Oct/0350.html

justin: correct, but no reason we cant

<WileyS> For those of us who actually need to implement we won't have the add'l bandwidth you're speaking to - only those that don't actually have to implement anything will have the bandwidth

justin: i hear args on both sides....

<WileyS> There will be a compliance regime in time for TPE testing

justin: we will move to TCS

chapell: my question is why so quickly?

<WileyS> David, the TPE has moved fairly swiftly when we focused just on it - don't you agree?

chapell: two reasons to wait... 1- bandwidth, 2. learnings can develop with release of TPE
... should i open an issue on this
... ?

justin: good quest.... i dont know
... i will talk with chairs after call on the process

<WileyS> That's because none of the chairs are working on real-world implementations

<dsinger> we're years behind where we thought we'd be; the group is showing clear signs of senility and tiredness. now we're working yes, there is light at the end of a long long tunnel. that's good.

justin: chairs all felt strongly made sense to move forward per opt 3 in the oct poll
... (missed>

chapell: i dont know i can say what the issues might be
... history points to that there will be significant learnings

justin: can you articulate these? you can time as we wont pick up until march

<WileyS> Once we begin injecting the DNT signal into backend systems we'll have a better perspective to understand what is and what is not possible from a compliance perspective

<WileyS> Not on the Server side...

<ninja> There is a complete process phase dedicated to addressing input from the implementation.

justin: .... we are going to test and there are imp now
... there may be competing compliance regime and we have made accomodations for those

chapell: to clarify
... position of chair is resp of members of working group that there will be potential learning after TPE released

justin: let's not over formalize this

chapell: whooo, i have concerns that the chairs are represnting the group on cases always and it might not be

(paaraphrased, correct if I got wrong)

schunter: (paraphrase) easier to do operability with both pie3ces
... i say we go ahead with TPE and see how it flies

chapell: should we be providing info now of potential things we might learn?

<WileyS> Matthias, you're hearing from actual implementors that we need time. As you're not an actual implementor, don't you believe you should listen to those of us who actual implement this standard in the real-world?

chapell: is there a change to postpone work on compliance spec based on the info

<JackHobaugh> It should be possible to test the TPE protocol without the TCS. Stubs can be used where necessary.

justin: speking for myself....
... it makes sense to discuss TCS and discussing those issues... but can be persuaded

<dsinger> I am confused at what I am hearing. For years people have complained that the W3C is unable to make progress, and now they are asking for delay?

justin: willing to honor will of the group

wendy: procedural, chairs have wide decretion on ordering and when to move forward...

<WileyS> One of the biggest issues we had with the TCS was detailed use cases to prove what was and was not possible in real-world systems. By giving us some time to implement the TPE (against "another" compliance regime) we'll have better data to give back to the working group for the more debated areas such as Permitted Uses.

wendy: heard from chairs they will listen to validated concerns

<Chapell> Wendy, you keep referencing the charter, but the charter doesn't specify timing

wendy: we thot getting both out for imp review and experience was important

<dsinger> http://www.w3.org/2011/tracking-protection/charter - see 2.2 Milestones

<wseltzer> Chapell, the charter has "milestones"

justin: shane you have thots on IRC. is your issue mainly about numbers

shane: ... biggest issues first time on TCS

<Chapell> Wseltzer, the charter also references tracking protection lists (:

shane: consumer adv wanted more real world data on why something not poss... and now we will be able to have real world data as evidence for why something will or wont work

<Chapell> Will we be doing tracking protection liss concurrently as well?

<dsinger> I don't see what has stopped that happening for at least a year; we have now changed the protocol in any important way in maybe a year now

shane: all these types of roadblocks that were proffered earlier can be provided now with ability to have real world use

<wseltzer> ["Progress of the Tracking Selection List specification along the recommendation track is independent of the progress of the two Tracking Preference specifications."]

shane: we will have diff compliance specs....
... that info will be critical to TCS conv to help with best path forward... and this will take time

<ninja> The official call for implementations comes with the formal status of Candidate Recommendation http://www.w3.org/2005/10/Process-20051014/tr.html#cfi

shane: those implementing wont have as much bndwidth

<sidstamm> dsinger, that's a good question.... why haven't we seen any preliminary data yet?

justin: your company has already been implementing against
... if you can bring this data forward that would help

<dsinger> Microsoft has gone ahead and implemented their half, so you even had a browser to work with.

shane: we have some... but we are refactoring due to how changes have come about
... but not completed... UGE, server side resp;onses
... still lots of imp and learning ahead
... don't want to be a slave...
... to the schedule

justin: i am not certain what they looks like or changes

<schunter> Let us discuss this more during the chairs call.

justin: ... yopu are prosping to wait some time after TPE and then TCS

shane: lets assume alt compliance is same time

<dsinger> what compliance side?

shane: want time to imp and learn and bring learning back to group
... lost opp if we just move forward right away

<WileyS> DSinger - MSFT still turns on the signal by default so that excludes them immediately

justin: you presummably have lots of data already

<dsinger> +q to Wendy to say that this is backwards. we do a draft, and then implement against that (mutually) and use that as a learning vehicle

justin: what arte you asking for 6 months, a year?

shane: no, a few months

<dsinger> WileyS, your opinion of the validity of the signal being sent in no way inhibits your ability to test against it

shane: should have more info(hopeful) end of q2 or q3 to bring info back

justin: hard to answer that going into this blindly
... i see args on both sides
... some say we already have delay
... this conv is helpful

<WileyS> David - disagree, how do I test UGE if I don't recognize the MSFT signal in the first place? It would require I honor it to test UGE. Paradox!

justin: can you put more of this in writing (i.e. the rational)
... chairs will have to talk

shane: who is pushing forward other than matthais

justin: maybe david aslo

wendy: we might be assuming more about the process given the name "l;ast call"
... group is trying to dev specs and learn friom imp process and change as needed fgrom learning
... and move forward

<Chapell> Wseltzer, appreciate your point, but learning from the implementation process will take bandwidth from those who are going to implement

wendy: we seperate TPE and TCS and both are needed, so makes sense to have both with basic spec that could be tested

<justin> ninja, you can scribe from this point forward?

wendy: no one is trying to rush

<ninja> yes

<ninja> scribenick: ninja

<kulick> (scribe was "lied to", told this was going to be a short call -- doh! ;)

<WileyS> Wendy - this misses the point that a W3C TCS may not be required and if it is, we'll have more information to inform that process with TPE implementations achieved

kulick: Back in the October poll, one of the arguments for separating was the chance to implement and gain more real-world information.

<wseltzer> [a diagram of the Rec-track process: http://www.w3.org/2014/Talks/chairs-part4/#/45 ]

<schunter> Should we move on and discuss this during our chairs call?

justin: Have not revisited the arguments in the poll back then. I think we heard some in this discussion.

kulick: Made the division assuming we had sufficient time to test.

justin: I will take a look at the arguments and the rationale of the moving forward decision.

<WileyS> David, That's what we're trying to do with the TPE

<WileyS> Intersting that the only people pushing back on the separation are non-implementors

dsinger: The Last Call is formally the status when we have agreed that we have a document we want to try to implement. Delaying the second spec would be taking this process backwards.

justin: We will do a 3 week break and giving people some time to think about/reviewing TPE.
... also time to calm down and think about how/when to take up Compliance.
... Thank you everyone for your hard work on TPE. Happy to have it in this almost Last Call status

<schunter> chairs call?

<dsinger> thx

<wseltzer> trackbot, end teleconf

rssagent, create minutes

thanks wseltzer

Summary of Action Items

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.138 (CVS log)
$Date: 2014-02-12 18:06:38 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.138  of Date: 2013-04-25 13:59:11  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Found ScribeNick: kulick
Found ScribeNick: kulick
Found ScribeNick: ninja
Inferring Scribes: kulick, ninja
Scribes: kulick, ninja
ScribeNicks: kulick, ninja
Default Present: hefferjr, Ninja, +1.858.229.aaaa, dsinger, JackHobaugh, WileyS, Fielding, eberkower, +1.323.253.aabb, kulick, Ari, justin, AWK, sidstamm, Wendy, MECallahan, schunter, Chris_Pedigo, SusanIsrael, Chapell, [FTC], LeeTien, [Apple], hober
Present: hefferjr Ninja +1.858.229.aaaa dsinger JackHobaugh WileyS Fielding eberkower +1.323.253.aabb kulick Ari justin AWK sidstamm Wendy MECallahan schunter Chris_Pedigo SusanIsrael Chapell [FTC] LeeTien [Apple] hober
Regrets: moneill2 robvaneijk CarlCargill npdoty
Found Date: 12 Feb 2014
Guessing minutes URL: http://www.w3.org/2014/02/12-dnt-minutes.html
People with action items: 

[End of scribe.perl diagnostic output]