W3C

Publishing Maintenance Working Group Annotation Task Force Telco

04 September 2025

Attendees

Present
ajellinek, Brady, dalerogers, gautierchomel, ivan, jim_saiya, LaurentLM
Regrets
-
Chair
LaurentLM
Scribe
gautierchomel, Brady

Meeting minutes

<LaurentLM> Use Cases and Requirements: https://w3c.github.io/epub-specs/wg-notes/annotations-ucr/

LaurentLM: welcoming Abe from Berkeley university, working with Zotero.

LaurentLM: We listed that we can annotate a text, an image and the part of an image

ivan: shall we add video?

ivan: and audio?

LaurentLM: that mean we have to tackle the time issue. I suggest to add audio only because video players will be more complex to get into to extract an exact position.

gautierchomel: I recently had an interaction on Vimeo where I added comment
… The interface made it easy to annotate a timestamp in the video
… But I agree let us stick with audio only

LaurentLM: the part of an image is interesting. Complex to do for developpers because that mean we have to draw.

LaurentLM: formatting annotations sounds a wish from users. Not sure how wide it is implemented yet.

ajellinek: we support that in zotero, we store annotations as HTML. standardising that across editor may be tricky. Not all editors accept HTML. Often we see subsets.

LaurentLM: maybe the note part can have a `content type` like HTML, MarkDown or PlainText, to drive the industry while letting space for flexibility.

LaurentLM: there are annotations and also bookmarks that need to be interoperable. That's part of the use case document too.

ivan: a general remark, what can be annotated, what can be bookmark? Can i annotate an annotation, a footnote? are there limitations to consider?

LaurentLM: this is not tracked yet. what does the group think about this?

gautierchomel: I think it is useful, but recognize it may add complexity
… there is a use case for annotating other annotations and footnotes

LaurentLM: footnote can come with difficulties to locate if they are outside the HTML.

ajellinek: we don't support annotating annotations,

ajellinek: but it is a reasonable use case

ivan: it is also related to where the annotation is stored. When coming to technical, we may not be able to implement.

LaurentLM: i see a way if we consider it a thread of annotations instead of annotating annotations.

ivan: it's different, there are use case where i want to select a part of an annotation to annotate.

LaurentLM: exporting & importing is a big part too. We have different scenarios here, from publishing a book with annotations to users sharing annotations, including in a classroom context.

DaleRogers: in the past authors responded books by publishing a new book, an article or a document. This technology may bring new use case as responding to an ebook with an annotation.

ivan: the use case I see is more in scholar or academic context, in a teacher to student relation or researcher communities.

LaurentLM: the implication of considering those use case is that we'll study how to standardise. We've hear about a known US author who sells annotated book and did that with popup because there was not standardised way.

gautierchomel: For annotation in social media, sharing annotations in a social media context

LaurentLM: I am an end user for this?

gautierchomel: Yes, I am in a book club, and I am sharing with other people in the club
… I could try to monetize that and sell annotations at 1 euro per book

LaurentLM: Can you write the text for that?

<ivan> +1 to gautierchomel

gautierchomel: yes, I will do that

LaurentLM: section 5 is about synchronizing.

LaurentLM: that mean among devices, among readers, among versions of ebooks.

ivan: I annotate the book i purchased, if i share with another person, they probably want to make sure they can differentiate the'y're annotations from mine. Identification is a use case to add here.

LaurentLM: among device is easy, it means I stand in my ecosytem, identification is easy. While sharing among users comes with identification complexities.

DaleRogers: i see it similar to customer support id tickets.

LaurentLM: unique identifier to ebooks is not sufficiently used

gautierchomel: Among device could also be among vendors
… so I read from one vendor, added annotations, now I want to use a new app
… so I have changed the ecosystem, and I should be able to do that

LaurentLM: So I can move annotations between reading systems

gautierchomel: I am not comfortable with the distinction between sections 4 & 5

LaurentLM: The big difference is the protocol
… In the first, it is just files, in the second there has to be some sort of protocol

gautierchomel: But many use case are in both
… So for a publisher providing them, they could be by file or protocol

ivan: back to the identification part, 5.1 is my ebook; 5.2 is other people ebooks. I have to refer to the book instead of my ebook.

gautierchomel: I think it lives in both

LaurentLM: it's true but use cases are to make sure we do not forget anything. Then file sharing and protocol may be used for all use cases.

LaurentLM: last point is how to make sure we survive versions. It's a real burden here.

ajellinek: we met the problem and still looking how to solve it.

<LaurentLM> Requirements for locators w3c/epub-specs#2763

LaurentLM: with the additions discussed today, I think we can start the requirement. An issue is open.

LaurentLM: we have no more time to discuss that today, I invite you to check and comment the issue.

Minutes manually created (not a transcript), formatted by scribe.perl version 244 (Thu Feb 27 01:23:09 2025 UTC).