23:48:40 <RRSAgent> RRSAgent has joined #epub
23:48:40 <RRSAgent> logging to https://www.w3.org/2021/09/16-epub-irc
23:48:42 <Zakim> RRSAgent, make logs Public
23:48:43 <Zakim> please title this meeting ("meeting: ..."), dauwhe
23:48:53 <dauwhe> Meeting: EPUB 3 Working Group Telecon
23:48:56 <dauwhe> Chair: dauwhe
23:50:14 <dauwhe> dauwhe has changed the topic to: https://lists.w3.org/Archives/Public/public-epub-wg/2021Sep/0015.html
23:51:31 <wendyreid> wendyreid has joined #epub
23:57:31 <shiestyle> shiestyle has joined #epub
23:57:45 <MasakazuKitahara> MasakazuKitahara has joined #epub
23:57:51 <MasakazuKitahara> present+
23:58:01 <wendyreid> present+
23:58:07 <dauwhe> present+
23:58:16 <toshiakikoike> toshiakikoike has joined #epub
23:59:24 <toshiakikoike> present+
00:00:30 <shiestyle> present+
00:01:03 <MattChan> MattChan has joined #epub
00:01:30 <duga> duga has joined #epub
00:01:32 <MattChan> scribe+
00:01:36 <mgarrish> present+
00:01:37 <dauwhe> present+ duga
00:01:38 <MattChan> present+
00:01:39 <duga> present+
00:03:14 <MattChan> TOPIC: PING Meeting
00:03:26 <dlazin> dlazin has joined #epub
00:03:35 <dlazin> present+
00:03:40 <MattChan> dauwhe: as part of horizontal review I worked on security and privacy questionnaire
00:03:56 <MattChan> ... requested review from PING, and they put us on their schedule
00:04:04 <MattChan> ... wendyreid and dauwhe joined that meeting
00:04:46 <MattChan> ... they expressed significant concerns in two areas. one is remote resources. Given that an ebook with remote resources then whoever hosts the remote resource knows i'm reading that book.
00:05:07 <MattChan> ... two concerns with scripting in general. Malicious script could send info back to a website if not blocked
00:05:24 <MattChan> ... on a more general level, the differences between epub and web were concerning to PING
00:05:29 <MattChan> ... no clear idea of origin
00:05:55 <MattChan> ... they feel users don't know that RS may be tracking behaviour
00:06:09 <MattChan> ... a lot of concern around user knowledge and expectations
00:06:27 <MattChan> ... supposedly web users are supposed to understand 1st party vs 3rd party, the nature of the URL, etc.
00:06:43 <MattChan> ... PING will have a bunch of notes for us, which will raise spec issues
00:06:58 <duga> q+
00:07:10 <MattChan> wendyreid: this doesn't need to happen today, but we have to discuss remote resources at the very least
00:07:17 <dauwhe> ack duga
00:07:17 <wendyreid> ack duga
00:07:38 <MattChan> duga: this is not surprising. we've not had a processing model, we've not described transports that we use for content
00:07:50 <dauwhe> q+
00:07:54 <MattChan> ... PING is look at this as if we are part of the web browser, or part of the web
00:08:09 <MattChan> ... but there's not reason that has to be the case. RS can be unconnected
00:08:24 <MattChan> ... so a lot of these concerns don't apply to epubs
00:08:48 <wendyreid> ack dauwhe
00:09:24 <wendyreid> q+
00:09:36 <dauwhe> ack wen
00:09:39 <MattChan> dauwhe: PING mentioned that there's no way for user to know that content has not been unadulterated since it came from the publisher. But in fact, you CAN be confident that RS has probably done a bunch of things to it
00:09:49 <MattChan> wendyreid: right, epub just doesn't fit into the same model
00:10:24 <duga> q+
00:10:49 <MattChan> ... in particular the thing with remote resources, e.g. Hachette could host a remote resource that is accessed by an epub, and the user would not know they are accessing a Hachette server
00:11:05 <dauwhe> ack duga
00:11:08 <MattChan> ... and that's fair, but most people also don't know who they're buying their books from
00:11:23 <MattChan> duga: and also, how is that different from the web? e.g. tracking pixels
00:12:26 <MattChan> dauwhe: wendyreid and I were trying to describe epub as it exists in today's world, but its true that we tend not to think along the lines of how could a malicious actor abuse this technology stack
00:12:55 <dauwhe> q?
00:12:58 <MattChan> ... we'll probably have to return to this topic on future calls
00:13:15 <MattChan> dlazin: what is PING expecting from us? Do we have to adhere to some policy by some date?
00:13:16 <mgarrish> q+
00:13:23 <dauwhe> ack mg
00:14:09 <MattChan> mgarrish: i remember going over some of this before, and we do say that there should be ways of alerting users to scripting and to disable it
00:14:21 <MattChan> ... but are we expected to begin prescribing UI?
00:14:50 <MattChan> wendyreid: we'll probably need to come up with some sort of solution (e.g. change to spec) that satisfies PING to the point where they are willing to sign off on it
00:15:01 <MattChan> ... just like a11y, or i18m horizontal review
00:15:22 <dauwhe> q?
00:15:26 <MattChan> ... they might say that they want to change something, and we can counter with a compromise
00:15:31 <dauwhe> https://github.com/w3c/epub-specs/issues/1793
00:15:32 <wendyreid> s/i18m/i18n/
00:15:48 <MattChan> TOPIC: Processing of Navigation Documents
00:16:08 <MattChan> dauwhe: issue is that nav is an HTML document, but also meant to be processed by RS
00:16:26 <MattChan> ... RS may use it and turn it into something unlike HTML, i.e. without markup
00:16:49 <MattChan> murata: I find that best practice is to use text only in nav
00:17:01 <MattChan> ... i am curious to know what happens when nav contains bidi?
00:17:07 <shiestyle> q+
00:17:09 <MattChan> ... will it be converted to bidi in unicode?
00:17:18 <dauwhe> ack shi
00:17:18 <MattChan> ... from bidi markup
00:17:22 <duga> q+
00:18:00 <MattChan> shiestyle: why does DAISY JP object to restricting rendering of nav?
00:18:16 <MattChan> murata: I don't want to restrict
00:18:37 <dauwhe> ack du
00:18:43 <MattChan> dauwhe: maybe just include a note to warn that html markup could break
00:19:24 <MattChan> duga: we hope that bidi is resolvable via unicode. We haven't gotten very many complaints, so if it breaks, its breaking in very few cases
00:19:39 <MattChan> murata: so those who create nav should make sure that their nav can be correctly rendered by major RS
00:20:10 <shiestyle> s/object to/want to/
00:20:11 <MattChan> duga: yes, always advisable. Some of these edge cases can be solved, but we haven't done it because no publisher has raised a strong issue about this
00:20:27 <MattChan> ... and we haven't heard much about ruby in toc
00:20:59 <MattChan> ... but maybe that's because JP publishers are tired of raising this issue
00:21:20 <MattChan> ... but if more publishers complain we will take that into account
00:21:48 <MattChan> ... for ruby we could put the ruby text in parentheses after the base text
00:22:08 <shiestyle> q+
00:22:15 <dauwhe> ack sh
00:23:22 <MattChan> shiestyle: many RS don't use rendering systems for toc, so publishers will not strongly request richly decorated tocs
00:23:35 <MattChan> ... i would leave the spec as it is for epub 3.3
00:24:23 <MattChan> dauwhe: i'm getting the sense that murata is okay with the change in language introduced by ivan
00:24:31 <MattChan> murata: yes, okay by me
00:24:33 <wendyreid> Proposed: Close issue 1793
00:24:36 <MAKOTO> MAKOTO has joined #epub
00:24:46 <wendyreid> +1
00:24:47 <dauwhe> +1
00:24:48 <MAKOTO> present+
00:24:49 <toshiakikoike> +1
00:24:50 <shiestyle> +1
00:24:50 <MattChan> +1
00:24:52 <MasakazuKitahara> +1
00:24:53 <MAKOTO> +1
00:24:53 <duga> +1
00:24:58 <wendyreid> RESOLVED: Close issue 1793
00:25:25 <dauwhe> https://github.com/w3c/epub-specs/issues/1782
00:25:31 <MattChan> TOPIC: Multiple Roles
00:26:11 <MattChan> dauwhe: about package metadata. someone proposed that we be able to assign multiple roles to a person using a comma separate string of different roles
00:26:27 <MattChan> ... but we don't think we want to get involved in parsing
00:26:46 <MattChan> ... but you can already assign multiple roles to a person in the current spec
00:27:03 <MattChan> ... and i'm not aware of many RS that even present this information to user
00:27:10 <dauwhe> q?
00:27:12 <wendyreid> Proposed: Close issue 1782 as won't fix. we
00:27:13 <mgarrish> q+
00:27:15 <MattChan> ... so i would close this issue
00:27:19 <dauwhe> ack mg
00:28:04 <MattChan> mgarrish: agreed. If we come up with a micro-syntax for the meta tag that could open up a bunch of issues. The one property per tag is a simple system that works.
00:28:05 <wendyreid> s/Proposed: Close issue 1782 as won't fix. we//
00:28:38 <wendyreid> Proposed: Close issue 1782 with maximum prejudice (won't fix)
00:28:43 <dauwhe> +1
00:28:46 <duga> +1
00:28:48 <shiestyle> +1
00:28:49 <mgarrish> +1
00:28:49 <MattChan> +1
00:28:49 <toshiakikoike> +1
00:28:50 <wendyreid> +1
00:29:00 <wendyreid> RESOLVED: Close issue 1782 with maximum prejudice (won't fix)
00:29:49 <MattChan> TOPIC: Can you TTS an image?
00:30:00 <MattChan> dauwhe: this is on the agenda, but its closed already
00:30:04 <MattChan> ... so we're good here
00:30:22 <MattChan> wendyreid: thanks to mgarrish's editorial wizardry
00:30:46 <MattChan> mgarrish: i was just waiting for some feedback from marisa
00:30:52 <MattChan> TOPIC: AOB?
00:31:10 <MattChan> dauwhe: okay, thank you everyone for this productive meeting
00:31:18 <MattChan> ... we'll see many of you next week
00:31:26 <dauwhe> Zakim, end meeting
00:31:26 <Zakim> As of this point the attendees have been MasakazuKitahara, wendyreid, dauwhe, toshiakikoike, shiestyle, mgarrish, duga, MattChan, dlazin, MAKOTO
00:31:28 <Zakim> RRSAgent, please draft minutes
00:31:28 <RRSAgent> I have made the request to generate https://www.w3.org/2021/09/16-epub-minutes.html Zakim
00:31:31 <Zakim> I am happy to have been of service, dauwhe; please remember to excuse RRSAgent.  Goodbye
00:31:36 <Zakim> Zakim has left #epub
00:32:46 <dauwhe> RRSAgent: draft minutes
00:32:46 <RRSAgent> I have made the request to generate https://www.w3.org/2021/09/16-epub-minutes.html dauwhe
00:33:01 <dauwhe> RRSAgent: make logs public
00:33:12 <dauwhe> RRSAgent: bye
00:33:12 <RRSAgent> I see no action items