EPUB 3 Working Group Virtual Locators Task Force Telco — Minutes

Date: 2021-04-14

See also the Agenda and the IRC Log


Present: Wendy Reid, Brady Duga, Dan Lazin, Marisa DeMeglio, Mary Coe

Regrets: Tzviya Siegman, Matt Garrish, Ben Schroeter

Guests: Pilar Wyman, Ronnie Seagren, Mary

Chair: Wendy Reid

Scribe(s): Brady Duga


Wendy Reid: https://docs.google.com/document/d/1y5JKcwlq1rvGYJ1HoLg0GF-NRiBsBNNZopGuQ0mgHUw/edit?ts=607779c8

Wendy Reid: overview - looking for a solution within epub for locators
… must work for content creators and RSes
… may need to change the spec, or may just create a note
… This is a hard problem, people have looked at it in the past
… Need to understand the problem space (specific use cases for locators)
… and what is the range of issues
… scholarly, trade, etc
… Asked for use cases, we got a lot back (in a Google doc)

Pilar Wyman: Don’t seem to be unique (lots of duplicates/overlaps)

Dan Lazin: Useful to cluster as unique problems
… probably no single answer that will hit everything
… grouping problems that might have the same or similar solution may help narrow the scope

Wendy Reid: Ok, let’s group them
… [listing each item]

Dan Lazin: Page numbers important for the first item
… but that may change over time when we replace page numbers
… but today we need page numbers

Wendy Reid: Case 2 - book clubs

Pilar Wyman: Good use case, very common if everyone is using ebooks

Wendy Reid: New group since this may not be page numbers
… Use 3 - deep linking from external source
… Falls under specific location

Wendy Reid: https://docs.google.com/document/d/1y5JKcwlq1rvGYJ1HoLg0GF-NRiBsBNNZopGuQ0mgHUw/edit?ts=607779c8#

Dan Lazin: Needs to be open to that location as well as display it
… Can you deep link now?

Brady Duga: Yes, epubcfi, not used

Wendy Reid: Case 4 - link to exact location, not artificial top of page
… brings up what to do about footnotes

Pilar Wyman: How do indexes work?

Wendy Reid: Reading index note from the doc being reviewed

Mary Coe: Do we want to talk about embedded indexing in depth now?

Dan Lazin: Maybe group first then come back to indexes

Ronnie Seagren: Is this a point locator (unlike the other uses)?

Wendy Reid: Putting in specific locations for now
… Case 5 - refer to a range of pages
… [reading note from the doc]

Mary Coe: People still like page numbers because they are used to them
… but once they look at the page number, they don’t care, or use the range

Ronnie Seagren: A range help tell you how much information you will find (ie it tells you size)

Brady Duga: Indicating size is something we haven’t considered before

Dan Lazin: There is an existing model, and we need to keep the information even if we change the model

Pilar Wyman: New standard drops page numbers, uses identifiers
… people seem to understand sections are large, subsections are small

Mary Coe: Need to understand what readers want

Ronnie Seagren: And there are many different kinds of readers

Dan Lazin: Users are used to ctrl-F (technical readers) so lots of nav not needed for technical web content
… books are different

Wendy Reid: Is this specific locations?

many: yes

Wendy Reid: Case 6 - RS wants to use real pages or screen pages
… seems to fall under page numbers

Dan Lazin: Is this the major implementation of page list

Wendy Reid: Maybe

Marisa DeMeglio: Thorium seems to do it
… but it has been a few years

Ronnie Seagren: This is also the case where a teacher tells students to go to a specific page

Wendy Reid: Case 7 - Reading the same content across different volumes
… Seems like specific locations

Dan Lazin: True, but it is actually throwing things away.
… That is it is doesn’t matter what the book is, it is relative to the start of content in the particular volume (anthology vs single book, etc)

Wendy Reid: Case 8 - Lawyer wants to reference some precise
… specific locations
… case 9 - auto generate locators
… Depending on where it is done (at creation vs reading time) will make it less general

Dan Lazin: If we put it in the spec how to generate it, then it can be consistent
… but authors would lose control

Wendy Reid: Page list is a problem for digital first/only
… Since there is no print version, there may not be a pagelist
… Is this specific location?
… Making a new category for this one
… Case 10 - cross language support
… translations may move the location
… it’s specific location

Dan Lazin: It is also ranges

Mary Coe: It is also less specific

Ronnie Seagren: May not be the same location as the sentence in the translated version

Dan Lazin: Seemed like arbitrary ranges, but paragraphs kind of solve this
… There are two versions to the problem

Wendy Reid: Putting it in specific
… Case 11 - annotations

Dan Lazin: Seems like specific locations

Ronnie Seagren: Once did an index where there could only be a single locator for every entry
… if there are multiple entries need a new sub entry
… nice because it was one to one (works well with links)

Wendy Reid: Continuously updated content, but locator that always takes you to the same place
… as another use case

Dan Lazin: pilarw2000__ mentioned poems, we should call that out explicitly

Pilar Wyman: Similar to translations, using line numbers helps

Wendy Reid: We have them all grouped
… might be all we can do today
… Looks like we have at least one large group, tackling that will hit most of the issues
… May want to concentrate on that (specific locations)
… might be a bad term

Mary Coe: Nope, I like it. But I guess we could have another name

Wendy Reid: Where to do from here?

Pilar Wyman: Also should look at the last one
… since software drives how we do things, it would be good to guide the software

Wendy Reid: No one handcodes their epubs
… lean toward authoring side for doing this, to avoid fragmentation of implementations

Pilar Wyman: When indexes are embedded you get an idea about the document and discussion
… that gets carried through to the digital side, even if they no longer makes sense (eg page numbers)

Mary Coe: Agreed. People like to see things like page numbers in the index, but once they are in the document they don’t care

Pilar Wyman: So I want this use case to stay
… this use case reflects how the software has changed how we think about this issue

Wendy Reid: Need an implementable solution

Ronnie Seagren: After we have a solution, I never want to see an epub without a linked index
… whatever we can do to make that a default

Wendy Reid: Not linked indexes are painful

Ronnie Seagren: Still need to convince publishers to add links

Mary Coe: Is there an RS that has an index button (similar to ToC, etc)

Wendy Reid: Index has never been elevated to same level as ToC in the epub metadata

Dan Lazin: Homework - a lot of what we have in specific locations can be solved with CFI
… Maybe we should go read CFI and figure out what would need to change to make it palatable
… make sense or is it prejudging the solution

Marisa DeMeglio: Real problem of use cases is “how do we stay webby” (ie solve this in the context of browsers)

Wendy Reid: HW - review CFI, think about where the gaps are, how will this work in a browser
… adjourned, meeting again next week at some other random time