13:56:59 <RRSAgent> RRSAgent has joined #epub-locators
13:56:59 <RRSAgent> logging to https://www.w3.org/2021/08/20-epub-locators-irc
13:57:01 <Zakim> RRSAgent, make logs Public
13:57:03 <Zakim> please title this meeting ("meeting: ..."), wendyreid
13:58:18 <wendyreid> meeting: Virtual Locators Task Force Telco August 20, 2021
13:58:29 <wendyreid> date: 08-20-2021
13:58:33 <wendyreid> chair: wendyreid
14:03:49 <duga> duga has joined #epub-locators
14:04:03 <ivan> present+ wendyreid, ivan, avneesh, dan
14:04:10 <duga> present+
14:04:25 <wendyreid> https://docs.google.com/document/d/11GypOjE9xOTaINATl5bxVIA3Mc9jzNBGCr6GT_KNaQ4/edit
14:06:05 <Hadrien> Hadrien has joined #epub-locators
14:06:12 <duga> scribe+
14:06:24 <avneeshsingh> avneeshsingh has joined #epub-locators
14:06:37 <avneeshsingh> present+
14:07:39 <ivan> present+ hadrien
14:07:45 <ivan> present+ will
14:08:34 <duga> wendyreid: Sorry for missing the last meeting
14:08:52 <dlazin> dlazin has joined #epub-locators
14:08:59 <duga> Dan: Most people that are here today weren't at the last meeting
14:09:30 <duga> wendyreid: Added a better definition of page list
14:10:03 <dlazin> Repost for those who missed the link: https://docs.google.com/document/d/11GypOjE9xOTaINATl5bxVIA3Mc9jzNBGCr6GT_KNaQ4/edit#heading=h.gbsmtjlkuon
14:10:10 <dlazin> present+
14:10:42 <duga> wendyreid: [Reviewing comment in the doc]
14:13:05 <Hadrien> q+
14:13:20 <wendyreid> ack Had
14:13:35 <duga> Hadrien: Is the page size calculated or authored?
14:13:48 <duga> ... if it is calculated than fine, we can use characters
14:14:05 <duga> ... but if it is authored then there can't be a fixed rule
14:14:32 <duga> ... calculated vs authored is an important distinction
14:14:35 <Will> Will has joined #epub-locators
14:15:05 <duga> wendyreid: The guiding principle is authors should provide their own list, but in the absence of that we have a calculated alternative
14:15:40 <duga> Hadrien: If we follow the terminology then page list should only refer to the authored content
14:15:59 <duga> ... so this definition is confusing
14:16:11 <duga> wendyreid: Agreed. Maybe just remove the fixed size reference here
14:19:28 <duga> Hadrien: I am concerned about using the same term ("page") for two different things
14:19:56 <duga> ... they are used in similar ways, but they aren't really the same, so using the same term is confusing
14:21:01 <duga> dlazin: So are you saying page is ok in the UI, but we should have different terms for authored page list page and the calculated page
14:21:25 <avneeshsingh> q+
14:21:38 <wendyreid> ack avneeshsingh
14:22:20 <duga> avneeshsingh: Would be good to have different terms
14:23:46 <duga> Hadrien: We can look at existing implementations and see what terms have been used
14:24:12 <duga> ... In readium they are called positions
14:25:18 <duga> q+
14:25:29 <duga> ... can have multiple positions on the page
14:25:40 <duga> ... used for a lot of things
14:25:55 <wendyreid> ack duga
14:26:12 <wendyreid> duga: We also have positions, and does the same thing. We don't use it when the page list is missing
14:26:28 <wendyreid> ... one issue is a page list, but you can also have a list of positions
14:26:42 <wendyreid> ... the way we handle a missing page list is to create a fake one
14:26:52 <wendyreid> ... it's based on the RMSDK
14:27:27 <wendyreid> ... we would create a not-great page list that would be consistent but not as good as an authored oen
14:29:07 <Hadrien> q+
14:31:28 <wendyreid> ack Hadrien
14:31:51 <duga> Hadrien: Calculated positions are in memory and not persisted
14:32:07 <Will> Will has left #epub-locators
14:32:11 <duga> ... But may persist for annotatiions
14:32:41 <Will_> Will_ has joined #epub-locators
14:33:12 <duga> ... I imagine page and position are complementary and both exist for different purposes
14:33:51 <duga> ... One is a non-exhaustive string, the other is exhaustive (no gaps)
14:34:02 <duga> ... There will be different use cases
14:35:08 <duga> wendyreid: Working on this terminology is important, given the discussion
14:35:39 <duga> ... Kobo uses position markers, but it is not at all related to page lists
14:36:23 <duga> ... Kobo is a locator, need a term for a calculated page
14:36:42 <ivan> q+
14:36:54 <wendyreid> ack ivan
14:37:34 <duga> ivan: Is the difference calculated or not?
14:37:50 <duga> ... Annotations, etc, are dynamic and to me that is a position
14:38:15 <duga> ... And the other is from the author and is set
14:39:09 <ivan> present+ tzviya
14:39:59 <wendyreid> https://docs.google.com/document/d/11GypOjE9xOTaINATl5bxVIA3Mc9jzNBGCr6GT_KNaQ4/edit#
14:41:02 <Hadrien> q+
14:41:06 <duga> wendyreid: Summarizing. There is page and calculated page, but we want a better name
14:41:14 <wendyreid> ack Hadrien
14:41:15 <duga> tzviya: What about calculated page?
14:42:15 <duga> Hadrien: Reviewing the doc at trying to decide if we are creating page lists or position lists
14:42:29 <duga> ... Some use case want page, but others work better with positions
14:42:47 <duga> ... So the answer to the question of what we really want to make isn't clear
14:42:53 <duga> ... it looks like we want both
14:43:17 <duga> ... For some, dealing with strings, it will be hard to implement a UI
14:43:54 <tzviya> q+
14:44:03 <wendyreid> ack tzviya
14:44:14 <duga> wendyreid: We don't consider UI much, we probably should consider it
14:44:33 <duga> tzviya: For some it won't matter if it is calculated or not, just so long as they are the same
14:44:36 <Hadrien> q+
14:45:01 <duga> ... The UI is an important point and is an issue we have had in the past with calculated pagination
14:45:15 <duga> ... Images are particularly a problem
14:45:27 <wendyreid> ack Hadrien
14:46:05 <duga> Hadrien: For use case 1, if there is a string and kids need to enter it they will make mistakes
14:46:16 <duga> ... due to typos, etc
14:46:52 <duga> ... For the second case, it does matter if it is authored or calculated
14:47:10 <duga> ... since you need a reference that works across print and epub
14:48:08 <duga> wendyreid: Authored and calculated will the terms and I will make an attempt at defining them
14:48:41 <duga> ... Maybe review the use cases as homework
14:48:52 <duga> ... And think about the desired end product
14:49:18 <duga> ... Any thoughts?
14:49:28 <duga> crickets: chirp
14:50:02 <duga> dlazin: Should note that a calculated page won't exist if there is an authored page (mutually exclusive)
14:50:18 <tzviya> q+ to ask about images etc
14:50:19 <duga> wendyreid: Maybe expand the section on persistence
14:50:39 <duga> dlazin: I just added that
14:50:42 <wendyreid> ack tzviy
14:50:42 <Zakim> tzviya, you wanted to ask about images etc
14:51:00 <ivan> q+
14:51:03 <duga> tzviya: Can we add images, math, etc into the process
14:51:04 <wendyreid> ack ivan
14:51:20 <duga> Hadrien: If you are using bytes then mathml just works by magic
14:51:48 <duga> ... Depends on what you are counting
14:51:55 <duga> ... images less so
14:52:13 <duga> ... requires rendering first
14:52:57 <duga> ivan: Going back to what Brady said, the calculated page list " bad" and is meant as a stick to convince authors to include them
14:53:12 <duga> ... So maybe we shouldn't spend too much time worrying about the perfect algorithm
14:53:46 <duga> ... This is an absolute fallback
14:54:03 <duga> Hadrien: There is quite a bit of complexity to manage the algorithm
14:54:53 <duga> ... When making the list you will likely have different information available than when you are actually displaying the page
14:56:34 <duga> all: general discussion of how hard it is to figure out the size of markup
14:56:52 <duga> dlazin: Added a section to the doc just now to cover this
14:57:06 <duga> ivan: What we are talking about here is for the static page list
14:57:16 <duga> ... It is not for annotations, etc. Is that correct?
14:57:29 <duga> Hadrien: You could enhance an annotation with that information
14:58:03 <duga> ... the simplest thing you can do for annotations is the original text
14:58:39 <duga> ... Then you can add position information to annotation like how far in the doc it is
14:58:44 <duga> ... which is good for anchoring
14:59:01 <duga> ivan: I think tha tbacks my claim, there are distinct things here
14:59:21 <duga> ... the static and dynamic are both useful
14:59:40 <duga> Hadrien: The one thing all reading systems need is the position list
15:00:12 <duga> ... Why do you need to calculated a page list for any use case
15:00:52 <duga> Hadrien: Why does GPB calculate the page list?
15:01:08 <ivan> rrsagent, draft minutes
15:01:08 <RRSAgent> I have made the request to generate https://www.w3.org/2021/08/20-epub-locators-minutes.html ivan
15:01:21 <ivan> zakim, end meeting
15:01:21 <Zakim> As of this point the attendees have been wendyreid, ivan, avneesh, dan, duga, avneeshsingh, hadrien, will, dlazin, tzviya
15:01:23 <Zakim> RRSAgent, please draft minutes
15:01:23 <RRSAgent> I have made the request to generate https://www.w3.org/2021/08/20-epub-locators-minutes.html Zakim
15:01:26 <Zakim> I am happy to have been of service, ivan; please remember to excuse RRSAgent.  Goodbye
15:01:30 <Zakim> Zakim has left #epub-locators
15:02:42 <ivan> rrsagent, bye
15:02:42 <RRSAgent> I see no action items