Meeting minutes
EU Text Data Mining Reservation protocol
EU TDM Reservation Protocol [draft]
Liisa: Laurent Lemeur
brought this to our attention
… Cristina noted this raises some concerns for
publishers and copyright
… what users of text data mining will do with content
… Laurent shared a draft paper with us
… thinking about where this impacts people's
businesses
Tzviya: I've been working on TDM at Wiley and in STM
<liisamk> link to TDM paper - https://docs.google.com/document/d/1NwtWv_ESS4ZhaHDHWnDSQSJKHrfRmLNjDTPdwNE3aeA/edit
Tzviya: STM is doing a lot
of work in conjunction with CrossRef
… so there's something embedded in the metadata to
make it clear what the license is and whether there's an exemption
… what a "scholarly article" is and what an
"exemption" means is not currently well-defined
… you need to allow the user to access the information
… and some sort of temporary storage is allowed
… this is of great interest to Wiley
Ralph: is this just an FYI from Laurent or is he proposing some sort of activity?
Liisa: Laurent would like
some feedback
… he had a conflict for today but hopefully we can
discuss in future meetings
Tzviya: I'll try to pull
together some stuff
… the solution is very silo'd
… if metadata can be embedded in some other identifier
then perhaps this can be a more universal solution
… the idea is to add metadata to an identifier
… if academic and trade publishers do this different
from scholarly, that will be a problem
Liisa: was anyone else
previously aware of this?
… I suspect there will be feelings about how and where
this gets used
… it will take some discussion
Ralph: who are the people in our community who we'd want to try to get to a future BG discussion?
Liisa: let our members know
this is going on
… and get them to provide feedback to Laurent
… do we think this is good or bad?
… you need metadata to follow along with the content
to allow or disallow
Tzviya: my impression is
that this is in legislation, like GDPR
… I don't know what exactly this will mean in the
world of trade publishing
… it will be a matter of coming to terms on how we
incorporate it into our workflows
… but I don't think we can opt-out
Dave: this feels out of scope for this group to me
Daihei: from the perspective
of Japanese publishers; in the next month or go we will go into deep
discussion on how we apply the new EU circumstances in the Japanese
industry
… I'll come back to you with some understanding of how
the Japanese publishing industry wants to apply TDM as well as more
generically the accessibility situation
ODRL Vocabulary [W3C Recommendation, 2018]
Ralph: perhaps an overview for BG participants on the use cases that the work is trying to solve
Liisa: yes; a use case
understanding that we can bring back to our internal teams
… companies need to figure this out for themselves but
awareness is our job here
Bill: the business issue is
not being surprised that your content is being mined without our
knowledge, or that you want it to be mined
… this sounds analogous to the EU accessibility
requirements
… the business issue is for publishers not to be
unpleasantly surprised
Wendy: this may fall into a
CG discussion; e.g. adding another metadata property to published
works
… we need use cases
Liisa: we'll ask Laurent to bring us use cases
Audiobook Recommendation status
Wendy: next steps: we're
still looking for implementors
… we got two more pull requests to be added as
implementors in the past 24 hours!
… we're putting together a list of content
implementations as well as implementations that are not official yet
… OBI is officially there as a content-producing
implementation
… I'll be adding Coresource; they plan to implement in
Q1 2021
… I'll be talking with Harper Collins about being
listed as a publisher implementor
… anyone else you know of who is willing to publicly
say they're planning to publish in Audiobook form, I'm happy to add
… we expect to start preparing for a Call for
Consensus to advance [to Proposed Recommendation] in September
… which will put us on track for Recommendation in
late November
MathML Refresh
Liisa: should we circulate this information?
MathML Refresh Community Group summary
WG draft charter [Neil Soiffer, 20-July]
Ralph: Brian Kardell from
Igalia gave a presentation on their MathML work at TPAC to the PWG
last year
… the MathML community group working on the refresh is
now discussing a charter for a working group
… Brian told us that the CG is identifying the subset
of MathML that is both implemented and used
… so this isn't redefining MathML, instead trimming it
down to make it easier to implement
<Avneesh> MathML WG also includes work on Chemistry
Ralph: I haven't had the chance to review the draft charter, but MathML is important to this community
<Bill_Kasdorf> There will be two versions, the core "trimmed down" version and the full untrimmed but improved version.
Ralph: this is something
that this community should provide feedback on
… bring them use cases
Ralph: as we know from the EPUB3 WG charter process, there's a lot of time between starting and shipping a charter
Liisa: Do you think this charter will improve MathML conformance in EPUB?
Liisa: do we feel there are still challenges in implementing and deploying MathML that will help the WG?
George: in addition to the
MathML CG there's a CG on chemistry
… the Chemistry CG has been working with the MathML CG
to add chemical semantics
… the first point was to identify the content as
chemistry
… and add metadata suggestions to allow
differentiation between symbols; e.g. 'K' can be potassium or Kelvin
… good work in the chemistry domain
Ivan: looking at the draft
charter, it has a goal of two specs:
… a simplified MathML Core and a MathML 4 that adds
terms and features
… it seems to be a big charter
… may require a lot of work to shepherd it through the
process
… they have a long road ahead
Wendy: from what Brian told
us at TPAC last year, the main challenge is not how MathML is
referenced
… but rather on the reading system side
… webkit doesn't support MathML natively
… we have to use a library to support it consistently
… and App Stores get angry if your package size is too
large
… if Igalia's work is adopted by enough browsers, this
will be a major change for us
… support in webkit alone will be great
<Bill_Kasdorf> +1 to Wendy
<JulieBlair> +1 Wendy
<Bill_Kasdorf> It's a big deal and it's really happening
Liisa: yes; the
implementation has always been the struggle
… there are some issues around authoring too
Wendy: yes; authoring has been a challenge but one of the biggest challenges was lack of interoperability
<Bill_Kasdorf> Would love to stay but have to drop for another meeting
Ralph: It's great that this is happening
George: Mellon Foundation provided the funding to NISO and NISO engaged Igalia
Tzviya: the integration with
CSS and Aria is one of the more important aspects
… the charter has so much in it; it might need to be
trimmed
… integrating MathML into Aria will be very useful and
important
Ivan: early feedback on the
CG's draft charter would be useful
… a clear standard for what is implemented to have a
core
… until that is done, extending beyond what's in
MathML 3 seems a recipe for problem
Avneesh: it seems that
Chrome may ship something within a year
… the charter has many deliverables; we need to help
them streamline
Ralph: we need to know what the publishing community feels should be streamlined
Use cases and content types
Liisa: we'd like a TF in the
CG for three content types:
… FXL + reflow
… what kinds of publications would benefit from being
able to combine these?
… Fullpage image + Reflow
… where the image could vary depending on the screen
… mix of FXL and Reflow in a single file
… so the reading system can select
… Daihei noted that this would be useful to the web
comics publishers, so people could see something as if it were
Reflow
… get a discussion going to see how these things might
be implemented
… should we start a document and start sharing stuff?
Ralph: Part of the function
of the BG is to find people who ought to be participating in these
conversations
… whatever it takes
Wendy: I struggle with this
… they are all already EPUB features and in the spec
… in the spine
… is this a CG thing to understand the use cases?
… so the spec properly communicates how to do it
… it's not as clear in the spec as it needs to be
… maybe a GitHub thread on use cases is enough; it
might not need to be a full Task Force
… we need to understand what people are trying to do
and how we are failing them
Liisa: I said "task force"
because that's what the CG chairs proposed the way the CG will work
… use cases are helpful; when there's a better
understanding that it works and how, it can come back to the BG to
help the reading systems understand how much content there will be
… putting together some use cases and handing them off
… showing what works, what works with some tweaking
… will help push the conversations
Avneesh: technically this is
in the standard
… but it would be good to have some sample content and
best practices
… people start implementing more easily with sample
code
Daihei: the issues here are
...
… whether it's a formal TF or not, we should explore
use cases and samples
… these subjects will help contribute to the
advancement of business
… expansion of business
… I would like to see Publishing@W3C help get such
contributions
… as you know, manga is a key element of digital
publishing in Japan; 85% of the business is manga
… webtoons is making a big impact on manga in Japan
… the form of FXL rendering is going to be a key issue
for digital publishing around the world
… the example cases could be really helpful
… and W3C's presence will be really appreciated
Tzviya: do we have a method for documenting these use cases?
Liisa: GitHub ?
Ralph: we can ask around the Team for patterns for gathering use cases
PBG Meetings during TPAC
Liisa: Daihei, Cristina, and
I would to propose that we have two BG meetings on the same day as
part of TPAC
… two 90-minute meetings on the 13th of October
… what big topics should we discuss there?
… should we propose cross-group meetings there?
Ivan: the EPUB WG needs to
plan its virtual F2F
… it's like to meet for several days
Wendy: I proposed some dates
TPAC Group Meetings scheduling wiki
Wendy: 19 to 23rd
<Avneesh> Oct 26 week should be week of breakout sessions
Ralph: we're asked to keep
the week of 12-16 October free for joint group meetings
… so the PBG meeting on the 13th isn't necessarily a
show-stopper
… but some of our participants and hoped-for observers
may have conficts with scheduled joint meetings
<tzviya> https://www.w3.org/wiki/TPAC/2020/GroupMeetings
Daihei: we'll come back with a proposed agenda
[adjourned]