13:05:41 <RRSAgent> RRSAgent has joined #pm-ann
13:05:46 <RRSAgent> logging to https://www.w3.org/2025/10/30-pm-ann-irc
13:05:46 <Zakim> RRSAgent, make logs Public
13:05:47 <Zakim> please title this meeting ("meeting: ..."), ivan
13:05:47 <ivan> Chair: laurentlm
13:05:47 <ivan> Meeting: Publishing Maintenance Working Group Annotation Task Force Telco
13:05:47 <ivan> Agenda: https://lists.w3.org/Archives/Public/public-pm-wg/2025Oct/0036.html
13:05:47 <ivan> ivan has changed the topic to: Meeting Details 2025-10-30: https://lists.w3.org/Archives/Public/public-pm-wg/2025Oct/0036.html
13:42:12 <ivan> regrets+ duga
14:01:54 <ivan> present+
14:02:06 <ivan> present+ toshiakikoike
14:02:18 <ivan> present+ kimberg
14:02:57 <ivan> present+ clapierre
14:03:11 <ivan> present+ ajellinek
14:06:05 <ivan> present+ hadrien
14:06:48 <ivan> present+ laurentlm
14:07:17 <Hadrien> Hadrien has joined #pm-ann
14:07:23 <ajellinek> ajellinek has joined #pm-ann
14:07:44 <CharlesL> CharlesL has joined #pm-ann
14:07:55 <CharlesL> present+
14:07:58 <CharlesL> scribe+
14:09:28 <LaurentLM> LaurentLM has joined #pm-ann
14:10:19 <CharlesL> LaurentLM: user requirements - use of "locator" its called in W3C it's called 'selector'
14:10:39 <CharlesL> ... fixed some renumbering. this is all I done in UCR.
14:10:58 <CharlesL> ... UCR document is sufficient to announce our profile to W3C annotations.
14:11:05 <CharlesL> ivan: do we want to publish the Note?
14:11:09 <CharlesL> LaurentLM: not yet.
14:11:35 <CharlesL> ... in UCR we want to say which requirement is a must have.
14:11:39 <Hadrien> q+
14:12:15 <CharlesL> LaurentLM: how do we write the profile, do we have help from Matt or not?  What help can we have for this writing.
14:13:10 <CharlesL> Hadrien: UCR: I have cold feet - last reading location, I understand the need but the work we are doing is not aligned with that one.  I think that is transient info.  I see interest in doing that but could see it done in other means.
14:14:15 <CharlesL> LaurentLM: DAISY ODP in DAISY world is a last reading location its at the cataloged protocol.
14:15:00 <Hadrien> https://apidocs.bookshare.org/reference/index.html#_highlights
14:15:05 <CharlesL> Hadrien: but Bookshare API does have this affordance.
14:15:14 <Hadrien> https://apidocs.bookshare.org/reference/index.html#_bookmarks
14:15:29 <CharlesL> LaurentLM: we can ask the DAISY folks, Avneesh.
14:15:32 <LaurentLM> q?
14:15:37 <CharlesL> ack Hadrien
14:15:37 <Hadrien> https://apidocs.bookshare.org/reference/index.html#_reading_activity
14:15:39 <Hadrien> q-
14:15:50 <CharlesL> ... Can we have help from Matt?
14:16:04 <CharlesL> ivan: depends on what help?
14:18:20 <CharlesL> ... Matt is potentially overloaded at this point and there may be ISO work for him to do as well.
14:18:43 <CharlesL> LaurentLM: Not worried about the pure writing of the spec per-say.
14:18:55 <CharlesL> ... the structure of the document is what is more concerning.
14:19:12 <CharlesL> ... we define the overall structure of the document.
14:19:37 <CharlesL> ... the def. of the structure would be helpful to get.
14:21:06 <CharlesL> Charles: Matt may have the boiler plate for this.
14:21:25 <CharlesL> LaurentLM: Its not really the boilerplate its more the structure of the document itself.
14:21:51 <CharlesL> ivan: we do have other notes which can be used as a start.
14:22:06 <CharlesL> LaurentLM: do we have a W3C template which defines a template of a profile?
14:22:13 <CharlesL> ivan: not recently.
14:23:00 <CharlesL> ... term "profile" is not a defined term, we can do what we want.  we should try to minimize the spec and try not to repeat things.
14:23:14 <CharlesL> LaurentLM: biggest part of the spec will be around 'selectors'
14:23:37 <CharlesL> ivan: overall structure is the same, these are the selectors we will use some will come from text others we will define locally.
14:23:48 <CharlesL> LaurentLM: this is why we will reuse the vocab of the original spec.
14:24:23 <CharlesL> ... # of entities annotation is a resource, its also a body, target and selector. So the vocab should be simple.
14:25:03 <CharlesL> ... what constitutes our profile, its very wide when we speak about annotations spec the body of the annotation could be 0..n can have several bodies.
14:25:13 <CharlesL> ... annotation can have 1 to many targets.
14:26:04 <CharlesL> ivan: one idea - follow the structure the annotation spec itself, in the annotation spec we talk about body etc. in our case its only 0 / 1. following the same structure of the toc level.  would help folks refer to the original spec.
14:26:53 <CharlesL> LaurentLM: map the spec to the use cases define some other terms, we speak about highlights, notes, etc. there is an easy mapping. if an annoation has no body this is a highlight or a bookmark.
14:27:14 <CharlesL> ... if an annotation has 1 body then we can say its a note.
14:27:28 <CharlesL> ... we have sub terms for note, highlight and bookmark.
14:27:58 <CharlesL> ... the content of the body is a note. is a type of annotation.
14:28:18 <CharlesL> ivan: when we write it down is where we will determine if this structure will work or not.
14:28:37 <CharlesL> ... What about JSON/LD encoding?
14:28:45 <CharlesL> LaurentLM: in a sense.
14:29:12 <CharlesL> ... the W3C spec JSON/LD can be very huge and w3c ann spec restricts/ subset.
14:29:20 <CharlesL> ... its really JSON lite.
14:29:44 <CharlesL> ivan: in this case maybe we should add a schema.
14:29:58 <CharlesL> LaurentLM: JSON schema would be good to add yes.
14:30:13 <CharlesL> ... something formal to add to the spec.
14:30:28 <CharlesL> ivan: in theory any JSON can be represented in Schema.
14:30:45 <CharlesL> Hadrien: I think you can do a broad schema for JSON/LD
14:30:58 <CharlesL> ... there are some reserve values.
14:31:49 <CharlesL> ivan: the JSON schema can define this thing.  JSON Schema was a moving target it should be more stable.
14:32:26 <CharlesL> LaurentLM: we can talk about the structure of what we will have.  A header, body and target.
14:32:47 <CharlesL> ... Header - w3c annotation header with the context
14:33:02 <CharlesL> ... its 1 context http...
14:33:08 <Hadrien> +1 for simpler context AND JSON Schema for our work
14:33:23 <CharlesL> ivan: the context file in annotation spec is very complex.
14:33:51 <CharlesL> LaurentLM: iri there is a type there is a creator modifier and modified.
14:34:25 <CharlesL> ... we can have a motivation in the header. one value is "bookmarking" should we define sub-types?  different sets?
14:34:28 <Hadrien> +1 for an enum of motivations in our spec and JSON Schema
14:34:57 <CharlesL> ... should we subtypes note, bookmark, highlight, maybe motivation types.
14:34:57 <Hadrien> q+
14:35:17 <CharlesL> ... w3c annotations no notion of notes / highlights.
14:35:25 <CharlesL> ivan: if it makes it easier thats fine.
14:35:40 <ajellinek> q+
14:35:46 <CharlesL> Hadrien: an ENUM would be important and in JSON schema. dev. only read JSON schema.
14:35:58 <ivan> ack Hadrien
14:36:29 <CharlesL> ... what is tuff, selecting text, highlighting it then bookmarking it. we may not be able to differentiate.
14:36:53 <CharlesL> ... small lists well documented.
14:37:18 <CharlesL> ajellinek: agree keep it small and let folks do what they want highlight / underlines
14:37:37 <CharlesL> ... I don't want separate motivations, in our models we will define which is which.
14:37:44 <CharlesL> ivan: motivation can be used for this.
14:38:34 <CharlesL> LaurentLM: user has selected something but didn't add a note so its a simply annotations. the color highlight properties is a broad term and doesn't describe the visualization but describes the structure.
14:38:39 <LaurentLM> q?
14:38:47 <CharlesL> ack ajellinek
14:38:49 <LaurentLM> ack ajellinek
14:38:49 <ivan> ack ajellinek
14:39:13 <CharlesL> LaurentLM: if we agree about the shape of the header, I can write something about that. then there is the target.
14:40:22 <CharlesL> ... multiformat target:URL, target:object with id, or source, many things in the target. choose if we want, but validated by schema. simple structure always a target or flexible object  or url, what properties should we keep?
14:40:56 <CharlesL> ... Readium target can have source and some metadata defined elsewhere. defined as simple.
14:41:17 <CharlesL> ... item in the spine and selector is the annotation for the source.
14:41:34 <CharlesL> ivan: closed world having a URL does not make a lot of sense.
14:42:02 <CharlesL> LaurentLM: said in req. we may want to have the target of one annotation can be another annotation.
14:42:51 <CharlesL> ... if we want to refer to another annotation we use the ID not source/selector
14:43:20 <CharlesL> ivan: how to define the URL is restricted to a URL to another annotations we don't want to allow a URL to my home page for example.
14:43:45 <CharlesL> .. we define annotations URL scheme for annotations themselves.
14:43:56 <CharlesL> LaurentLM: UUID is used in Readium.
14:44:27 <CharlesL> ... 4 blocks with hyphen in the middle.
14:44:47 <CharlesL> ... defining the selector will be a little tricky
14:45:20 <CharlesL> ... choose the formats of the text, might be HTML / markup if its not a text image video audio, we will need a URL which points to what?
14:45:30 <CharlesL> ivan: do we want version 1 to do that?
14:45:33 <CharlesL> LaurentLM: No.
14:45:59 <CharlesL> ivan: HTML fragments is doable. but an annotation to a video could be difficult.
14:46:13 <CharlesL> LaurentLM: overall structure we just need to know the pointer type.
14:46:38 <CharlesL> ivan: EPUB if its a video then video must be part of the EPUB package.
14:47:05 <CharlesL> LaurentLM: for audio annotations its my voice recorded as the annotation.
14:47:17 <CharlesL> ... deaf I use sign language as the video annotation
14:47:33 <CharlesL> ivan: that means adding a new file to the EPUB.
14:48:03 <CharlesL> ajellinek: it would be unlikely to add that in our system.
14:48:19 <Hadrien> q+
14:48:30 <CharlesL> ... image annotations would be difficult.
14:48:39 <CharlesL> LaurentLM: we remove it from the use case?
14:48:45 <CharlesL> ivan: use cases are OK.
14:49:20 <CharlesL> LaurentLM:  we don't remove from UC, but if the annotation is on the web and can b a URL.
14:50:06 <CharlesL> ivan: by talking a reader can conversion from audio to text the audio doesn't have to be stored potentially.
14:50:21 <ivan> ack Hadrien
14:50:43 <ajellinek> q+
14:50:57 <CharlesL> Hadrien: there are ecosystems that specialize in that. kindle scribe. in their case working with image is the most natural thing. we are missing representation, but maybe they are not interested in interoperability.
14:51:07 <CharlesL> ... there are an increasing # of those devices.
14:52:34 <Hadrien> q+
14:52:49 <ivan> ack ajellinek
14:52:59 <CharlesL> ajellinek: distinction image and drawing on a screen, we would like to allow for ink annotations - relation to the text, circle something positioning to reflow able content is not really possible. especially when screen size changes. vs. attaching a image, but ink annotations are much smaller.
14:53:02 <ivan> ack Hadrien
14:53:32 <CharlesL> Hadrien: depends on how you store it, context fore reflowable depends on user settings font size spacing etc.
14:53:59 <CharlesL> ... both the annotation and the text being annotated and stored together as an image.  this is one way to get it out of a specialized ecosystem.
14:54:08 <CharlesL> ... something to consider.
14:54:51 <CharlesL> ajellinek: 1/2 solution we would like to show something you draw on the page, but when you resize things not sure how we would show something on the page in a seamless way. maybe as a popup.
14:55:06 <CharlesL> ivan: didn't Edge do this before switching to chromium.
14:55:33 <CharlesL> Hadrien: I know some of them doesn't do well with reflowable content for anchored annotations.
14:56:02 <CharlesL> LaurentLM: I can start writing something. maybe before TPAC.
14:56:16 <CharlesL> ... maybe file structure.
14:56:34 <CharlesL> ivan: I won't be there. too late for me at that time.
14:56:48 <CharlesL> Brady will be around and Hadrien.
14:57:08 <CharlesL> ajellinek: I hope to be there virtually.
14:57:16 <CharlesL> Charles will be remote hope to be there.
14:57:52 <CharlesL> LaurentLM: We will look at this and the w3c annotation spec during TPAC.
14:58:56 <CharlesL> ivan: TF are freely managed by the WG.
14:59:16 <CharlesL> Hadrien: if we feel we need a new TF we can create one.
14:59:25 <CharlesL> ivan: yes the Chairs will need to set that up.
15:00:02 <CharlesL> CharlesL has left #pm-ann
15:01:05 <ivan> rrsagent, draft minutes
15:01:06 <RRSAgent> I have made the request to generate https://www.w3.org/2025/10/30-pm-ann-minutes.html ivan
17:02:42 <Zakim> Zakim has left #pm-ann
18:52:39 <LaurentLM> LaurentLM has joined #pm-ann