Web Annotation Working Group Teleconference

<trackbot> Date: 05 August 2015

<trackbot> Meeting: Web Annotation Working Group Teleconference

<trackbot> Date: 05 August 2015

<shepazu> https://lists.w3.org/Archives/Public/public-annotation/2015Aug/0012.html

<fjh> topic agenda https://lists.w3.org/Archives/Public/public-annotation/2015Aug/0012.html

Agenda Review, Scribe Selection, Announcements

<fjh> ScribeNick: RayD

26 aug cancelled

<fjh> RESOLUTION: 26 August meeting cancelled

RESOLUTION: 26 August meeting cancelled

<fjh> String Matching and Searching draft is now available, http://www.w3.org/TR/2015/WD-charmod-norm-20150721/

Ivan.string matching. doug looked at it

<fjh> Note that RangeFinder is joint deliverable with WebApps so we should plan of CfC for FPWD in both groups as well as heads up re timing

<fjh> https://lists.w3.org/Archives/Public/public-annotation/2015Jul/0210.html

Minutes Approval

doug: think we should do a cfc, need to fix algorithms
... case-sensitive matching recommended. described how to do unicode casesensitivity

<fjh> proposed RESOLUTION: Minutes from 22 July approved, http://www.w3.org/2015/07/22-annotation-minutes.html

RESOLUTION: Minutes from 22 July approved, http://www.w3.org/2015/07/22-annotation-minutes.html

TPAC

<fjh> reopen questionnaire? https://www.w3.org/2002/09/wbs/73180/annotation-tpac_2015/results

<bigbluehat> +1

there was a questionaire, not many responses, should we reopen?

<Kyrce> *shepazu, fjh no worries. But thanks, it is true that I get a lot of creative pronunciations!

re-open it, need to decide if there will be a meeting

(I won't be going)

Rob: september at the latest.

ivan: warning, airfares are increasing don't wait till last minute

<shepazu> now open until 12 September

rob: what number of people do we need?

<fjh> All, please complete questionaire

Ivan: more important, will we have the people we need

<fjh> please note that this is NOT registration for TPAC, you need to register for TPAC separately

doug no point in a meeting without at least 3 implementers

rob's going

paolo not going

<fjh> http://www.w3.org/2015/10/TPAC/

Don't know if Nick's going

doug: could be having a "testing" meeting

ivan: what about social web guys?
... cross pollination

fjh: is the deadline passed?

ivan: think we said we'd have a meeting

doug: they don't take kindly to cancellations

<bigbluehat> fjh: yep. I'll be there. buying tickets this week (hopefully)

fjh: think it will be productive, given the people who already said they'll be there

<bigbluehat> +1 for having dial-in

fjh: does dial in make sense?

rob: probably

doug: skype or IRC or webex makes more sense than dialin

<bigbluehat> the Q1 questionnaire is linked above

question about dates

<bigbluehat> what's the URL of the correct one?

<fjh> I will send out email reminder

<bigbluehat> got it. tnx

Data Model

<fjh> Hypthes.is and JSON - https://lists.w3.org/Archives/Public/public-annotation/2015Jul/0194.html

bigbluehat: annotator doesn't think open wordly, doesn't store enough
... hypothesis does better
... don't know where to put XPointer, maybe as a fragment id
... be good if could get examples, scenarios, maybe an annotation primer

rob: model documents should defend why a particular feature is there

<fjh> Thanks takeshi for noting the needed correction to the questionnaire

<bigbluehat> +1 to getting the JSON-LD context doc at a URL soon

<azaroth> +1 to both :)

tim (Not hearing tim clearly)

<fjh> s/tim.*/tim: need context doc soon

bigbluehat: thought experiment, native json, like if wordpress started using json
... got valuable stuff but not ready for prime time. plan to reformulate the json
... would be useful to keep track (on wiki) of the different annotation formats used

<Jacob> maybe put the context document on next week's agenda?

<fjh> s/tim (Not hearing tim clearly)/tim: need context doc soon/

context document for json-ld

<ivan> www.w3.org/ns/anno.json

ivan; practice is to put it on w3c namespace

tim: if it changes is that a problem?

ivan: no

tim: useful if it was there, for work on validation

<TimCole> no, I don't expect to be at TPAC...

rob: should it have all the protocol related matters?

ivan: our choice
... does it have to include the ontology?

rob: will do a preliminary draft today

<ivan> anno.jsonld

<azaroth> http://www.w3.org/ns/anno.jsonld (yes?)

multiple bodies

rob: another proposal.
... interested in the process by which we come to a decision

ivan: if you don't want to do anything with roles you can ignore it. Makes it more attractive.

<azaroth> https://lists.w3.org/Archives/Public/public-annotation/2015Aug/0036.html

paolo; like proposal

<azaroth> +1 to not requiring reference to annotation

don't need reference to annotation

<Jacob> +1 and +1 for employing Ray's solution

scribe: ugly in json
... approach requires some interpretation, but it's local, and you can ignore it

<bigbluehat> PaoloCiccarese: (or anyone) do we have an example of the "stitched up" JSON? :)

tim: like proposal. concerned that those on list but not on call will see this as unnecessary indirection
... need concrete examples.

<Jacob> Is it possible to discuss the pros and cons of Ray's approach compared to the use of hasScope that Lutz's annosys example uses? IIRC, it was another possible solution.

<Jacob> +1 for a wiki page cataloging the pros / cons of each approach

<bigbluehat> +1 to the wiki page

<tbdinesh> +1 for the wiki page

<bigbluehat> hard to know clearly what we're discussing without something to look at :)

<Kyrce> +1 to the wiki page

<PaoloCiccarese> +1

tim: tim will put up a page for this

<Zakim> azaroth, you wanted to note framing complication

rob need to be careful not to embed body into role assignment

ivan before we do that wiki page (to compare things) need to be sure we have at least one

<azaroth> +1 to Ivan

scribe: example that has all the predicates correct, and both json and turtle

<Jacob> +1

<bigbluehat> +1

scribe: example in turtle can become complicated in json-ld

<azaroth> My explanation of why we don't need the link to the annotation: The introduction of the resource solves the open world issue. It can be used by multiple annotations that have the same resource with the same role.

doug -- not clear on all this and that's why want a wiki page, to help understand

<fjh> +1 to doug, we need clarification here.

paolo - not sure a wiki page is going to help explain.

ivan start with a clear proposal, with the predicates etc.

<azaroth> +1 to PaoloCiccarese

paolo - constraints from being based on sematic web concept. Not everyone cares about semantic web

<bigbluehat> +1

scribe: and we need to explain why we care?

<Zakim> azaroth, you wanted to ask if role == motivation

<Jacob> +1 to Paolo

rob is the role a motivation, or not. propose that it is.

tim concern about motivatedBy with multiple targets, can't do it

<fjh> s/tim concern/TimCole: concern/

scribe: what's appropriate for body might not be appropriate for target

<fjh> s/rob is/azaroth: is/

<Jacob> basically a role is a sub-class of motivation

rob ok with multiple predicates, but nervous about not using Motivation class.

paolo different functions. but ok with range the same

Other Business

<fjh> All, please note questionnaire is corrected, so please take a look and answer it https://www.w3.org/2002/09/wbs/73180/annotation-tpac_2015/

scribe: might be confusing to some but has to be clear that range is the same

<fjh> no call 26 Aug, reminder

<azaroth> Proposed action on the group was to use a new predicate for the relationship between assignment resource and the role, but to use the Motivation class as the range to avoid duplicating all of the instances

doug - maybe better if people answered the tpac poll

<azaroth> And as per Ivan to create examples in both Turtle and JSON-LD to see how it looks

Adjourn

<ivan> trackbot, end telcon