W3C

Web Annotation Working Group Teleconference

08 Oct 2014

Agenda

See also: IRC log

Attendees

Present
Benjamin_Young, Dan_Whaley, Dave_Cramer, Frederick_Hirsch, Kyrce, Kyrce_Swenson, Matt_Haas, Paolo_Ciccarese, Ray_Denenberg, Rob_Sanderson, TimCole, Tim_Cole, [IPcaller], azaroth, bigbluehat, chillag, dauwhe, dwhly, fjh, raphael, rayd, tilgovi, unknown, Jacob_Jett, Randall_Leeds, Jake_Hartnell, Raphael_Troncy
Regrets
Ivan_Herman, Doug_Schepers, Bill_Kasdorf, Ben_De_Meester, Luc_Moreau, Antonio_Olmo_Titos(tripu), Nick_Stenning, Markus_Gylling
Chair
Frederick_Hirsch, Rob_Sanderson
Scribe
rayd

Contents


<trackbot> Date: 08 October 2014

Scribe Selection, Announcements, Minutes Approval

<bigbluehat> can I map my name to bigbluehat?

<fjh> you can map your number to a name using this tool, it can help https://www.w3.org/1998/12/bridge/info/name.php3

<bigbluehat> awesome. I'll try

<tilgovi> Randall Leeds

<fjh> ScribeNick: rayd

no announcements

<fjh> proposed RESOLUTION: 1 October 2014 minutes approved with Rob added as chair: http://www.w3.org/2014/10/01-annotation-minutes.html

minutes accepted with addition that rob is chair

RESOLUTION: 1 October 2014 minutes approved with Rob added as chair: http://www.w3.org/2014/10/01-annotation-minutes.html

Extended Data Model Introduction

Paolo walk through second half data model

<fjh> http://www.slideshare.net/paolociccarese/open-annotation-specifiers-and-specific-resources-tutorial

paolo: used to think about core extensions but now single namespace

… specifiers and specific resources slide 2

<fjh> slide 2

scribe: slide 3 basic anno model. can annotate one or more targets and have one or more bodies
... body and target can be URI
... but most of time want to annotate a fragment of a resource
... uri fragments have limitations so we need to get beyond that
... ... slide 5 fragment of an image
... same for body
... called specific resouce not specific body or target
... slide 7
... tool to specify selectors

<fjh> slide 8 notes types of specifiers, including styles

scribe: highlight with different colors to indicate different things
... URI one way to indicate a fragment, but they don't cover all the use cases
... (URI fragment)

<raphael_> Raphael: The Media Fragment URI specification, http://www.w3.org/TR/media-frags/, specifies how to specify simple temporal fragments (using two time codes) and simple spatial fragments (reduced to rectangle bounding box using a top,left position + width and height, aka xywh)

scribe: some selectors built in but you can make up your own

<fjh> paolo: spec has a number of selectors that have been generally agreed, could define additional selectors but need interoperability

scribe: 10
... 12 list of builtin selectors
... range, etc.

<fjh> paolo: fragment URIs had some disagreement but other representations had URI issues, but so far seems the best choice

scribe: 13
... 15 fragment selector

<fjh> slide 14, source + fragment, e.g source1#namedSection

scribe: 16 breaks URI into source and selector, conforms to media fragment spec
... 17 SVG selector to identify fragment of image
... ... can define squares, circles, ovals

<Zakim> fjh, you wanted to ask degree of concern about use of fragment URIs

<fjh> raphael: ok with fragment URIs but also need to be able to create structured data

<fjh> raphael: concerned about validation

raphael need to make sure that what you're breaking down is consistent with URI and not clear how

<fjh> ISSUE: How to validate fragment selectors

<trackbot> Created ISSUE-7 - How to validate fragment selectors. Please complete additional details at <http://www.w3.org/annotation/track/issues/7/edit>.

raphael: media spec allows simple fragment, rectangles etc.
... need use cases for more complex cases

rob: some were offered at CG

<fjh> summary so far: we are ok with media fragment URIs for simple cases of time and box region selection, need use cases to justify more complicated cases

<fjh> yet paolo notes for biology there are more precise annotation selection cases, also for maps

paolo: there are plenty of cases
... with SVG seletor can do many of these things

<raphael_> Raphael: the question is then whether complex regions (e.g. arbitrary complex polygons) should actually be defined in the hash part of a URI or not

paolo: but there are other ways to do it
... maphub uses svg as well as text selector
... they think text selector easier to encode

<raphael_> Raphael: maphub uses svg and WKT

<fjh> fjh: is there an issue to record

<fjh> paolo: no, extensibility takes care of it

paolo: selectors are important but take care to avoid proliferation

tcole: potential for fallbacks in selectors

<fjh> http://maphub.github.io/

paolo: some selectors only specific applications understand but others can fallback

<fjh> paolo: should keep this in mind, can have complicated selector and also simpler one for fallback (eg shaped region or rectangle)

rob: sequence of fallbacks rather than random choice

<fjh> azaroth: IDPF had ePub need for this

rob: issue 2 in github

paolo: 19 text selectors. Issues, copyright for example, counting characters
... text quote v. text position
... 21 a paragraph is annotated and there is text before and after

<azaroth> Or if the change is big blocks of text being moved around

<azaroth> And the prefix + exact + suffix moves together

paolo: 22 text position selector, probably simplest.
... 23 start and end

<fjh> paolo: need to be clear text quote and position depend on document stability.

<fjh> csillag: have implemented both of these, sometimes you might want to use multiple methods at once

paolo ... 24 state

scribe: keep track of when page was annotated
... time and http request state are recorded
... 25 time state

<fjh> paolo: state was discussed as “setup”, e.g. rotate 3D model so it is same view as when annotated

scribe: 27 http request state
... selectors can be used together with state to properly identify target
... 29 style specifiers
... can say annotation is styled by css style
... 31 scope specifier. Might annotate an image w/in a web page

<fjh> paolo: issue raised of whether annotation is for image in context of page, or independent of page. My experience is that context generally matters.

scribe: e.g. scope is web page and source is image in the web page

<fjh> paolo: pragmatically annotate the image and then note the scope

scribe: 32 annotated telescop image

tim: selector state straightforward but how to implement use cases.

<fjh> tcole: concerned about use cases supporting style, scope and maybe should revisit

tim: style and scope specifiers need more work

rob: hope we don't need to revisit whether there is a requirement for them, but discuss how to implement

paolo: tim right, we never reached agreement

rob: will generate two issues.

<azaroth> ACTION: azaroth to generate "status quo" proposal issues for Style and Scope [recorded in http://www.w3.org/2014/10/08-annotation-minutes.html#action01]

<trackbot> Created ACTION-1 - Generate "status quo" proposal issues for style and scope [on Robert Sanderson - due 2014-10-15].

<fjh> azaroth: please review issues in github and discuss on list

Other business

<fjh> none

Adjourn

Summary of Action Items

[NEW] ACTION: azaroth to generate "status quo" proposal issues for Style and Scope [recorded in http://www.w3.org/2014/10/08-annotation-minutes.html#action01]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.135 (CVS log)
$Date: 2009-03-02 03:52:20 $