EPUB 3 Working Group Virtual Locators — Minutes

Date: 2021-05-19

See also the Agenda and the IRC Log

Attendees

Present: Pilar Wyman, Brady Duga, Ben Schroeter, Ronnie Seagren, Mary Coe, Dan Lazin

Regrets:

Guests:

Chair: Wendy Reid

Scribe(s): Brady Duga, Wendy Reid

Content:

Wendy Reid: https://docs.google.com/document/d/1y5JKcwlq1rvGYJ1HoLg0GF-NRiBsBNNZopGuQ0mgHUw/edit?usp=sharing

Wendy Reid: Summary of last week
… George said let’s just use page numbers
… But there are questions around pagination in reading systems
… May not be possible to have consistent numbers
… So let’s pick it back up now

Ben Schroeter: Would it make sense to look at the use cases and see what would be handled by page numbers?
… For instance, ones with a start and end requirement

Pilar Wyman: Wants to put in a plug for page numbers
… The issue isn’t really consistent pagination, that happens already (different sized pages)
… we understand that for a given volume the numbers are consistent internally
… The real issue is direct to digital
… The only place we need to say there must be some pagination is for those non printed books
… May not be that big of an issue since we already aren’t consistent

Ben Schroeter: If the publication doesn’t have page numbers, we want to avoid virtual page numbers from reading systems
… since we want reading system agnostic page numbers

Pilar Wyman: I meant authoring not reading system

Brady Duga: I think George’s comment was that we should have a virtual page numbering system that is easy to do without pages
… google play’s secret, we respect the page numbers when provided, or use bytes
… page 75 is the same regardless of device
… there’s issues for citations on certain devices, encoding changes, might be imprecise
… might need to be unicode code points instead
… it’s doable and easy for both device, authoring system
… it feels like a hack
… but maybe it’s ok

Dan Lazin: Was going to suggest what we already do
… a few points: page numbers are already arbitrary
… If you have a print book, use it, otherwise we will make something up
… May be a problem when a digital first appears in print, or various formats
… We don’t have to have a single solution, can use pages, tick things off, then adopt other solutions when needed
… Maybe we should recommend page lists when at all possible
… Then suggest a fallback algorithm
… which is entirely arbitrary

Ben Schroeter: If we don’t do it at authoring and rely on reading systems
… we run the risk of non-identical page numbering

Dan Lazin: Algorithms should be dead simple
… aim for simplicity

Pilar Wyman: Want to add page numbers are not the only solution
… There could also be other numbering systems like section numbers
… The types of content that this will be necessary is fiction

Dan Lazin: Why is fiction different?

Pilar Wyman: Fiction has fewer breakouts
… Non-fiction has more sectioning

Ben Schroeter: A lot of fiction has no section

Pilar Wyman: Some of our use cases are fiction (or similar)

Brady Duga: Few points, page lists and non-device oriented page numbering can be fraught due to device ideologies, some just prefer certain page numberings
… differences across devices etc
… there’s entrenched opinions
… we’ll face a challenge convincing some people
… we might be able to couch it in supplying references/indexes, better ui

Dan Lazin: How does this role out? In the spec?

Wendy Reid: This would probably an extension of the navigation document
… If we have some fallback algorithm, then we would have to specify
… And this recommendation would have to go through the whole group
… and it would be at risk since we need two implementations

Dan Lazin: Is there some other model?

Wendy Reid: We could publish a separate document as a note
… And see how people react
… But we could try to get two implementations if we are confident

Ben Schroeter: Matt, what are your thoughts around note vs spec

Matt Garrish: Not really. Whichever you want to do is fine
… might want to start as a note. Easier to go note to spec than vice versa

Ben Schroeter: There was a perception that not having these citations was a barrier to adoption of epub

Ronnie Seagren: Don’t want this to go the way of the indexing standard
… want this to get adopted
… Do we have contacts

Brady Duga: We have done notes that have become specs, not an absurd way to go
… in terms of other reading systems, we have the other bigs, unsure of adoption
… in terms of people, we have play, kobo, others
… many of them are also w3c members

Wendy Reid: We have a lot of reading systems vendors listening

Matt Garrish: Some of this will depend on how strongly worded it is in the spec
… Could be in the spec but not a must

Wendy Reid: Best route is note, lay out the use cases and proposals
… There is probably still a CFI role
… Note first is also easier to distribute early
… Note is way easier to read than the minutes
… That means going back to the use cases
… need to narrow down to the core cases
… for instance, deep linking might not make the cut
… Then figure out which ones would work with page numbers, and which need something else (eg CFI)

Ronnie Seagren: What about getting some tools (eg InDesign) to adopt this?

Wendy Reid: We can always ask

Wendy Reid: https://docs.google.com/spreadsheets/d/1KO-HyLGUUw36F-ruAARHNiPO1aUJCCNeTv3zxGtjuHw/edit#gid=0

Wendy Reid: Added a new column to see if page numbers satisfy the requirement (in the linked spreadsheet)

Ben Schroeter: Do we also want a primary/secondary column?

Wendy Reid: Yes
… [Reading first entry]
… [continuing to read through, updating as we go]

Dan Lazin: Can we get a catchy name for the different types of page numbers
… to make it clear

Wendy Reid: Mary, your research came up. How do people feel about generated page numbers?

Mary Coe: I give them arbitrary page numbers in the index, and people seem to like them
… they are very familiar
… Pages, ranges, etc mean something, once they are in the text they no longer care
… But once in the text they use other markers (eg section numbers)

Ben Schroeter: virtual folio (catchy name suggestion)

Ronnie Seagren: How about the word “page link” [?]

Mary Coe: For now need more testing to see what people think

Dan Lazin: Looking at a reader app
… Looks like they have changed their page numbering recently
… Only shows current page when there is a page list

Brady Duga: What happens if you change the font size?

Dan Lazin: Consistent if there is a page list, but not when there is no page list
… Maybe chunks as a name? Or maybe “screens” vs “pages”? But people may not like that

Wendy Reid: Maybe the name will come to us
… [back to going through the spreadsheet]

Dan Lazin: Are page level bookmarks good enough?

Wendy Reid: Most reading systems are more granular
… But maybe RS to RS is good enough?

Dan Lazin: How about metric page numbers?
… decimal page numbers solve all our cases
… so subdivide pages into smaller sections

Ben Schroeter: so 1.027 or something?

Mary Coe: Metric might work for some cases (eg annotations or bookmarks)
… But when giving them a new location requires orientation
… Not always relevant

Dan Lazin: Could suggest displaying other context when or perhaps recommend ignoring decimal portion when creating eg an index

Ben Schroeter: Same issue with physical books and bookmarks

Ronnie Seagren: A rough location can be better, since the reader needs to understand the context around the reference

Dan Lazin: If I had an indexing system that links somewhere, I would want to go to the page and highlight the specific text

Pilar Wyman: When we embed indexes I link exactly where the target is
… sometimes you can’t even see the top of the page

Wendy Reid: Time is up! Homework - can we mark up the remaining use cases?
… If you disagree just add a comment and we can discuss
… My homework will be to set up a note (just a skeleton)
… Do people prefer html or markdown?

Dan Lazin: Why use html? Markdown is better

Wendy Reid: Ok, will use markdown