Meeting minutes
tzviya: apologies for the late agenda
Summary of PBG meeting
liisamk: there is concern
from the Japanese community
… that they want assurances about backward
compatibility
<Ralph> 14-April BG meeting record
liisamk: and they are
worried about requirements for implementations of EPUB features
… but there aren't concerns about specific features
<Ralph> Strengthen compatibility requirements, involve epubcheck #8
Daihei: we want to secure
assurances that throughout rec track compatibility will be assured,
and maintenance of the eco
… system
… we want it confirmed and written
Daihei: and new features
which might cause disruption
… and so backward compatibility will not be assured
… and there were concerns expressed about the HTML
serialization of HTML5
… which could affect implementation
ivan: I don't remember the
exact dates
… there were a series of issues coming into the
charter
… from Florian, from Makoto, from Toshiaki
… and some of the things you mentioned, were taken up
… and we had three or four pull requests that were
accepted and merged
… so the charter has been updated
… for HTML5 we had some minor changes
… the question came up about ssml or epub:type, things
that are bound to the xhtml serialization
… and we added words to make it clear that no RS can
refuse to take XHTML
<Ralph> current draft proposed EPUB 3 WG charter
ivan: that was settled
… the thing that becomes more complicated is
… the standard w3c approach to CR is that there need
to be two implementations of each feature
<ivan> https://github.com/w3c/epub-3-wg-charter/issues/19
ivan: I'll put the reference
to the issue here
… what happens if there's a feature with only one
implementation, therefore it doesn't meet CR exit criteria
… and they said it should not be removed from the
standard because there are documents that use it
… it is perfectly possible that if a feature doesn't
get 2 implementations
… it could be kept as a non-normative feature because
there is no interop
… but documents could remain conforming
… this came up again this morning
… with a new comment in the repo
… one thing I raised
… is that this process issue, is not required to be in
the charter
… because it's part of w3c process
… but we could make it explicit in the charter
… the charter makes very strong statements about
compatibility
… I'm not sure what else we can do to alleviate the
concerns
Ralph: thanks Daihei for
reinforcing the depth of the sensitivity
… what would help me to understand the types of
sensitivities
… things like changing version ID caused issues in the
past, so we promised not to do that
… A high-level concern is that products, that is
books, that are currently being shipped
… publishers don't want anything to happen to flag
those existing books as no longer proper
… they don
… 't want the business chain to throw out existing
inventory
… or not to be able to sell to a consumer existing
inventory
… how to express that in charter language
… Ivan has expressed technically how we might
implement that
… how we might implement tests for features without
interoperability
… is the kind of language that talks about the
production chain, is that kind of language something that is needed?
… there is a sensitivity to something we haven't yet
captured in language
dauwhe: I'm going to ask are
there any changes to EPUB that are acceptable?
… explicitly, new features
… having some new feature in EPUB, if you wanted a
reading system to conform, it might require development
… are we saying that there can be no change to RSs,
platforms, process
… I'm trying to find the boundaries of this discussion
… My goal is making the tent bigger, all content
remains valid but there are new possibilities
… are we objecting to new possibilities
tzviya: that's a good question
George: I'm asking the same
kind of question
… if we added html serialization, and still allow
xhtml
… the reading systems that would be able to render
HTML could be the two implementations
… if we allow that, is that OK? books could have html
or xhtml?
… and reading systems would support both, or maybe
just xhtml
wendyreid: I'm on the same
wavelength
… I responded to Daihei's issue on github
… I thought that the charter line was clear enough
… we intend to be backwards compatible
… and I thought we explained in the call we want to
make the tent bigger
… we want content creators to produce better epubs by
using current web tech
… but all books will be valid
… no one wants to invalidate the entirety of the
catalog
… what language needs to be changed? that's the
feedback we're looking for
jeff: I'm hearing a lot of
discussion about the tech of backward compat and the charter language
… we've made clear we're demanding backward compat
… but I'm still hearing concerns
… but maybe the answer is not in tuning language in
charter
… but having a set of conversations to give people
more confidence in the process
… the w3c process is new and scary :)
… in w3c the best way to get what you want is not by
negotiating charter language but by participation
… encouraging greater participation would be good
… there would be no break in compat wihtout consensus,
and no consensus if the people are participating
… so we need more conversation, maybe a third chair
from Japan
<Zakim> tzviya, you wanted to discuss process
tzviya: thanks jeff
… I agree with what Jeff said
… I think there's been misunderstandings about the
process
… I think the charter language is clear
… maybe the next BG should go over the process
… we're not talking about EPUBCheck, and the
validation process is not in scope for this group
… we decide what we mean by compatibility
… we can decide how this works
garth: I have such a long
list of +1
… I don't agree that if a feature is not supported by
any reading system it should be kept
… but we don't want to invalidate any content
… but if a feature exists in an EPUB, there was some
implementation that created that feature
… although that's a legalistic argument :)
… but I agree with Jeff that actual engagement will be
more useful
liisamk: it's not the people
who are unaware of w3c process, it's Makoto who has raised this the
most
… to Dave's point on new features, I think new
features are acceptable
… a line of assurance that the business chains, the
clarity that all those things are possible implementations
… one line of those things would help
ivan: what is that line?
… we have tried very hard to get the charter text to
address these concerns
… we need something like "this line is wrong, change
it to X"
liisamk: we can talk about the process in the next BG call, but that doesn't. help us today
Avneesh: It's a trust issue.
Some people said they are not part of the WG.
… there is not enough participation in the WG from
them, so it can slips out of their hands.
<Daihei> I have been on q+ and waiting
<liisamk> +1 to Avneesh
Avneesh: we could have
invited experts from Japan, if people are not members
… we also said that CG will be point of innovation
… why not have the incubation in the CG, even if we
know it won't break anything
… let the CG build confidence before moving something
to the WG
… re: epubcheck
… EPUBCheck should not be the reference point for
compat
… we have relaxed EPUBcheck things so we don't break
the long tail of old content
<wendyreid> +1 to Avneesh
<Zakim> Ralph, you wanted to comment on "one line"
Ralph: I hear ivan's
frustration
… that we need precise language
… I would be willing to sit with Liisa and Daihei
… I know that Florian who knows process well
… Florian offered 2 very long sentences
… if it's a question of trust, then being involved in
the group is the way to be confident about that trust
… if we can make the charter address that more
explicitly, it's worth the effort
… if liisamk and Daihei are willing, I can take an
action
wendyreid: I agree with this
… ??? brought up the issue of not being able to join
the WG
<liisamk> +1 to help with the "sentence"
wendyreid: we should
encourage people to become w3c members to participate in the WG
… it's clear they want to participate. We need them
there.
… even with the minutes and meeting summaries, it's
hard to communicate all the information in a way that's
understandable to everyone all the time
… some Japanese publishers are full members, and we
should encourage them to join
… encourage the Business members to participate as
full members, and to join the CG
… I like the idea of a Japanese co-chair, but we need
someone who has the technical skills and wants to chair
Adaret: coo
tzviya: I agree with Wendy
… there are probably more Japanese publishers who are
full members than US publishers
<Ralph> [I have had experience with chairs whose expertise is in group facilitation rather than the technical details of the deliverable. Those were very successful]
tzviya: we would love to have their technical people on the WG
<Zakim> tzviya, you wanted to discuss ie for wg
tzviya: it's about people
being able to participate
… if Daihei or FLorian have suggestions, we would love
to hear from them
liisamk: Wendy, it's hard in
BG to encourage people to become full members
… joining the W3C is not easy financially
… especially right now
… we talked to Christina the other day. Publishing is
down to nothing there.
… we don't want to discourage existing members, but we
can't push people to do more. It's not a reasonable ask right now.
Daihei: agree with liisamk
… we would like to have the participation
aderet: squeak
Daihei: of Japanese members,
but there is a financial issues
… I will be sure to discuss with next week all the
Japanese w3c members
… to appoint someone as a co-chair, and I will get
back
… I know that participation from japan is important
… Kodansha, Shueisha, MediaDO, etc
… and other companies considering becoming full member
… In terms of the language
… the business chain or ecosystem
… that is what we wanted in
… but if that's not technically good enough to be in
the charter
… I will check with the technical people for the
language
… I don't think that Japan is the only one who wants
backward compatibility
… and existing files and products and inventory to be
used in the future
… last year, there was an epubcheck error about the
ordering of content in TOC
<liisamk> +1 to Daihei- I can attest that lots of NA and European publishers want backwards compatibility
Daihei: people were blaming
me, as co-chair of PBG, that such an error message came up
… because they would have to spend millions of dollars
to fix the issue
… and that is similar to the other countries
… no one said no to new features. We do want to
improve epub!
… as long as we don't break anything
… I will have someone from Japan to participate in WG
… even outside of WG we should have democratic process
… that's what we need
jeff: several people have
commented about searching for a co-chair from Japan who has technical
depth
… I would suggest that we relax that requirement
… with multiple chairs, each of them have their own
skills and bring different things to the group
… if we had a third person in facilitating the input
from Japan's technical community
… that could be perfectly fine co-chair
Ralph: I've had experience
as team contact of exactly the scenario Jeff suggests
… with someone who was skilled in facilitating the
group
<Avneesh> +1 to focus on participation firs
tzviya: we need to focus on people who can participate, and we can think about chairing later
George: I'm concerned about
two things
… one is... what would disrupt the Japanese publishing
industry
… if new titles came into the market that were html
and had alternative HTML nav doc
… and publications start entering the market
… then existing reading systems would have to support
the new content
… would this disrupt the ecosystem?
… I see that as natural evolution and development of
standards
… but would it anger people?
… two: greater communication between the CG and the WG
is important
Daihei: in terms of
candidate for co-chair
… I have a couple people in mind
… but they might not work for w3c members
… so I would ask for consideration as invited expert
… to be able to join the WG
… to answer George's question
… I have a new feature but will not break
compatibility
… if the feature might expand business
… and work as a benefit to the RS provider
… so it wouldn't be seen as a negative
<Zakim> Ralph, you wanted to ask about reading systems
Daihei: improvement of epub will be welcomed by Japan and everywhere else
Ralph: thanks Daihei
… do existing reading systems have to support EPUB 3.X
content? That's a plausible tech constraint.
… it may be hard for the WG to weigh that kind of
constraint.
… how will existing reading systems behave wrt an EPUB
3.X doc?
ivan: two things
… one: to what George said, the way the charter reads
today
… on new features like HTML, there is a sentence which
said the effects of this feature must be carefully considered.,..
.... ....
… that makes it clear there's no obligation to add
html, just as there are no obligation around other possible new
features
… it's great we got these comments from Japan
… but we didn't get any comments from anywhere else in
the world
<wendyreid>+1
ivan: I feel uncomfortable with the charter that we have no comments from Europe. N America, Africa, China, Antartica, etc
Daihei: about epub reading
systems and new features...
… that is my opinion, I will come back to you after
confirming with japan
… the language for the world for the business chain, I
will come back to you
… and the co-chair candidate, I will come back to you
liisamk: PBG can put out an email to comment on the charter
Other GitHub open issues on the draft EPUB 3 WG charter
<Ralph> #2
<Avneesh> EPUB 3 community is in CG, so it is also good to remind PCG/EEPUB3CG
Ralph: issue #2, IDPF
registries
… it seems obvious to me that the w3c registries
should be part of the new WG
… is there a reason not to say that
<garth> +1
dauwhe: W3C should
definitely take responsibility for the registries
… as many of them as possible become part of the specs
RS Accessibility bugs - where to file?
Ralph: ivan, you can add that
liisamk: I talked to my team
about the PBG repo for RS bugs.
… should we include a11y bugs?
dauwhe: I think we should
absolutely include them
… they are bugs
… we can use labels to sort and triage
… having them all in one place is useful
… and people don't have to judge where their problem
lives
tzviya: if it's really about the book or the system or the AT, there is an ARIA WG task force, and they are working with JAWS etc on board
liisamk: there's a bug in ibooks if alt text is embedded in figure it's not available
George: I agree with Dave
… separate but equal is inherently unequal
<wendyreid>+1
George: triage is important
… I filed a bug with JAWS on that issue
Ralph: there is consensus :)
adaret: waves
Ralph: thanks everyone, and we'll do better on timely agendas