See also: IRC log
<trackbot> Date: 30 July 2014
<scribe> scribenick: npdoty
justin: short call
... next week we want to take up the market research / audience measurement
... authors comfortable with their existing proposal, take up that discussion starting next week
... the week of August 10th, no call as at least all the chairs are on vacation
... only a handful of issues still to deal with. will send around another potential batch closing email
... will send around, just let us know if you don't want any of those closed
<trackbot> issue-203 -- Use of "tracking" in third-party compliance -- open
justin: discussed previously.
fielding working on a proposal for restructuring
... should have that for us next week
<wseltzer> npdoty: I made the changes to the compliance document
<wseltzer> ... when you are tracking/not tracking, qualifiers, examples
<justin> npdoty: I've updated the spec to include TSVs along with examples in the different subsections.
justin: look at what npdoty has done, though expect it doesn't address fielding's issue
justin: have a consolidated
proposal from walter, mike and nick
... document is currently silent
... less specific to url shorteners exactly
... any questions or concerns about that language?
... unless there are any friendly amendments
... would go to a Call for Objections on that tonight
... feel free to send comments to the list, if any adjustments should be made
justin: sadly, we don't have
dsinger on the call today. not sure yet about how to merge with
... two seem firm. article 29 proposal from vincent; expert review/safe harbor from jack
... not sure yet about merging david and roy
... maybe that would also satisfy proposers from EFF (heard from Lee)
<fielding> maybe definitions for identified, de-identified, re-identified, and anonymous?
<vincent> is it similar to the green/yellow/red states?
npdoty: suggestion of having two terms, "deidentified" and "aggregate", where deidentified requires the three-step contract requirement but aggregate doesn't because it's so far from being identifiable, like "50%"
<WileyS> "Linked" is debatable in the Yellow State
justin: is there a similarity to
previous red/yellow/green proposals about which can be released
and which wouldn't
... test for "deidentified" (in dsinger proposal or current text) already. even aggregate numbers can enable reidentification
moneill2: if you delete the data and only return the count of scottish people, then there should be no reidentification concern
justin: just saying "aggregate"
doesn't solve the problem (like small groups)
... would be interested in the exact terms of the text proposal for this super-de-identified state
<WileyS> We could have an entire working group on deidentification - we're going too deep here. Not that its not needed, but too large an issue to solve here.
justin: encourage people to talk more on the list and develop one or two more consolidated proposals
<WileyS> I would recommend we stay high-level (like Roy's proposal) and drill down into the details in another forum/pass/working group.
<fielding> I have not seen any example given where the data could be re-identified after it meets my proposed definition, since that would imply it does not meet my definition. ;-)
justin: per Shane, this can be a very complicated area. argument might be for a simpler description in our text
justin: personalization proposals
last summer, perhaps related to lack of confidence in the
signal issue, to imply that some personalization is allowed
even when a DNT signal is present
... is there a proposal to include a "trackless advertising" term in the document?
... if that's the intent, it would be good to clarify these proposals
... there was a lot of discussion about that last summer [issue-215]
npdoty: is the intent to just clarify editorially where restrictions are present? or a change from what the current text currently prohibits?
justin: it might be that 236 is that sort of editorial thing, and that 234 is the intent for removing personalization prohibition
JackHobaugh: yeah, that sounds right
justin: will send email, don't need edits yet
justin: decision about data
append was announced, but text needs to be sent around before
... quick call. anything else to discuss?
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: npdoty Inferring Scribes: npdoty Default Present: Jack_Hobaugh, hefferjr, Fielding, Carl_Cargill, Wendy, eberkower, WaltMichel, npdoty, justin, Chris_Pedigo, RichardWeaver, vinay, kulick, WileyS, +1.323.253.aaaa, adrianba, moneill2, robsherman, [FTC], vincent, Brooks, walter, Chapell Present: Jack_Hobaugh hefferjr Fielding Carl_Cargill Wendy eberkower WaltMichel npdoty justin Chris_Pedigo RichardWeaver vinay kulick WileyS +1.323.253.aaaa adrianba moneill2 robsherman [FTC] vincent Brooks walter Chapell Regrets: johnsimpson sidstamm chris_m dsinger schunter Found Date: 30 Jul 2014 Guessing minutes URL: http://www.w3.org/2014/07/30-dnt-minutes.html People with action items:[End of scribe.perl diagnostic output]