– DRAFT –
(MEETING TITLE)

Me neither

<axelpolleres> Maybe than easier if Bert chairs...

Axel, could you try headphones?

<axelpolleres> ...apologies, have no headphones close by at the moment.

Agenda item 2: Approval of F2F minutes

<axelpolleres> PROPOSED: approve F2F minutes as per agenda links

<axelpolleres> +1

<harsh> +1

+1

+1

<Fajar> +1

Resolved: approve F2F minutes as per agenda links

<axelpolleres> harch sent a long list of action items we can close...

<harsh> mail: https://‌lists.w3.org/‌Archives/‌Public/‌public-dpvcg/‌2019May/‌0001.html

Will close the action items from Harsh's Email offline

Action: bert to close actions to be closed as per https://‌lists.w3.org/‌Archives/‌Public/‌public-dpvcg/‌2019May/‌0001.html

<trackbot> 'bert' is an ambiguous username. Please try a different identifier, such as family name or username (e.g., bbos, bertv).

https://‌docs.google.com/‌spreadsheets/‌d/‌13d1eRXZZBCw84vYGoCJeMU08rzkkzadDzxY3n2iOi8k/‌edit#gid=1356512815

Action: bbos to close actions to be closed as per https://‌lists.w3.org/‌Archives/‌Public/‌public-dpvcg/‌2019May/‌0001.html

<trackbot> Created ACTION-103 - Close actions to be closed as per https://‌lists.w3.org/‌archives/‌public/‌public-dpvcg/‌2019may/‌0001.html [on Bert Bos - due 2019-05-14].

Vocabulary spreadsheet

<axelpolleres> it should all be on the spreadsheet.

<axelpolleres> I essentially added all contributors as authors to the draft.

<harsh> https://‌github.com/‌dpvcg/‌extract-sheets/

Harsh tried to run the script to generate the documentation from the spreadsheet

<harsh> https://‌dpvcg.github.io/‌extract-sheets/‌index.html

<axelpolleres> there is a script now to generate a spec doc from the gdoc and the gsheet.

<axelpolleres> FWIW, I also had worked further on the overall document...

<axelpolleres> https://‌docs.google.com/‌document/‌d/‌1Z3Eb5rZjrdWcE5u5o0CYzA_LPyGaTqmg84ecGve_ZLA/‌edit#

<axelpolleres> Processing:

<harsh> Processing tab - not sure about description for (32) Exchange and (42) Restrict

Exchange is not in the GDPR, Restrict is

Rigo: Exchange is confusing - would need to list all kinds of disclosures and have a new class for all constraints - would be overkill

Rigo: we need transfer, but we don't need exchange.

Exchange = 2xtransfer

<harsh> Rigo: propose to remove exchange from Processing

<harsh> +1

<axelpolleres> question: do we need otherwise conditional disclosure?

<axelpolleres> aren't these very close to ODRL actions?

Axel: Could we not refer to ODRL ACTIONS for processing conditions?

<axelpolleres> I consider this an ISSUE (which we do not have to resolve now.

<axelpolleres> i.e.: ISSUE: do we need a generic mechanism to describe conditions and restrcitions to a specific processing.

<axelpolleres> PROPOSED issue: ISSUE: do we need a generic mechanism to describe conditions and restrcitions to a specific processing or can we re-use ODRL'

<axelpolleres> s mechanism here?

Rigo: Constraint is a very general concept; we need to find out what the best place is to put constraints in the taxonomy.

PROPOSED: make the modeling of constraints an ISSUE?

<axelpolleres> PROPOSED issue: ISSUE: do we need a generic mechanism to describe conditions, constraints and restrcitions to a specific processing or data category can we re-use ODRL's mechanism?

<harsh> +1

+1

<axelpolleres> +1

<rigo> +1

Resolved: raise ISSUEdo we need a generic mechanism to describe conditions, constraints and restrcitions to a specific processing or data category or can we re-use ODRL's mechanism?

Issue: do we need a generic mechanism to describe conditions, constraints and restrcitions to a specific processing or data category or can we re-use ODRL's mechanism?

<trackbot> Created ISSUE-19 - Do we need a generic mechanism to describe conditions, constraints and restrcitions to a specific processing or data category or can we re-use odrl's mechanism?. Please complete additional details at <https://‌www.w3.org/‌community/‌dpvcg/‌track/‌issues/‌19/‌edit>.

Axel: How do we link the issues to the draft document?

Just as links to issues?

Action: Harsh to discuss with Axel to make sure ISSUEs are reflected in FPWD

<trackbot> Created ACTION-104 - Discuss with axel to make sure issues are reflected in fpwd [on Harshvardhan Pandit - due 2019-05-14].

<axelpolleres> harsh: processing action "exchange" was removed (unclear what it meains

<axelpolleres> harsh: we should make sure that classes from the base ontology aren't dublicated in the subsheets.

Top-level concepts (dpv:processing etc.) should be put in the base ontology.

Action: harsh to remove duplicate top-level classes from subsheets.

<trackbot> Created ACTION-105 - Remove duplicate top-level classes from subsheets. [on Harshvardhan Pandit - due 2019-05-14].

timeline

Harsh regarding timeline: by the next call, there can be a complete draft version of the document to review.

<axelpolleres> harsh: By next call: spec should be finished. we will also try to make progresss on the document

<rigo> +1 sounds good

<axelpolleres> harsh: aims to send around document by this weekend.

<harsh> collaboration/help/assists welcome in the Google Doc

<axelpolleres> should we action some people for internal review?

<axelpolleres> or next time?

harsh: Should agree on how exactly to publish it.

<harsh> ACTION-99

<trackbot> ACTION-99: Bert Bos to Look into where to publish our cg spec, and how to redirect from the namespace doc to the spec. -- due 2019-04-12 -- OPEN

<axelpolleres> we need a vanilla HTML version of the doc ...

<harsh> by vanilla - you mean without Google Doc css/styles?

<axelpolleres> yes, it should use W3C spec stylesheets

<axelpolleres> or no?

<harsh> Yes, that would be better

<harsh> I'll try to strip out the CSS in the script itself, and then include w3c CSS

<axelpolleres> the trouble is at the moment we use gdoces instead of plain HTML, but we can maybe move to plain HTML with the FPWD and then stick with the plain HTML for editiing, leaving the google spreadsheets.

<harsh> we can generate vanilla HTML from gdocs as well by using a css-remover on the gdocs html

<rigo> the ad industry has 27000 classes for profiling, where is the limit?

<harsh> @rigo is there any way to access these 27k classes? Would be great to build into a vocabulary.

<axelpolleres> We can always refer to external vocabularies, if available, but we will not invent them

<harsh> Yes, I meant creating an external vocabulary instead of merging into dpvcg

<axelpolleres> We can discuss to name them as an extension, is there a link to these?

<axelpolleres> isn't that already possible? i.e. someone can mark data both as locationdata and as sensitivedata or just one of those?

<axelpolleres> I need to run, sorry.

If we include levels of sensitivity in the taxonomy, then that's some real knowledge we can transfer.

<axelpolleres> (We should adjourn to the next call in two weeks, FWIW or raise an ISSUE)

Issue: how to express sensity of data?

<trackbot> Created ISSUE-20 - How to express sensity of data?. Please complete additional details at <https://‌www.w3.org/‌community/‌dpvcg/‌track/‌issues/‌20/‌edit>.

<harsh> Propose: Derived data as a class instead of a property

Action: Elmar to change isderived to a class

<trackbot> Created ACTION-106 - Change isderived to a class [on Elmar Kiesling - due 2019-05-14].

<harsh> next meeting 23rd May

<Mark> Privmas Eve

<Mark> May 23rd -