See also: IRC log
See also the agenda of the meeting, with references to talks and slides.
thierry: (introducing workshop committee)
… thanks to Marcus and Angela for all hard work and organization
… we have also 19 experts in program committee
… thanks for their hard work reviewing all papers
… finally thanks W3C colleagues
… for experience, advice ; Karen for all logistics aspects and Maria at the registration/admin
… a few words about submissions
… we got 43 submission papers that covered many topics
… presentation, metadata, DRM, etc
… we reviewed all
… and we built agenda from there
… unable to accomodate everyone because too many papers submitted
… great interest in all the papers but need to have a subset to fit the 1.5 day schedule
… we have 90 attendees representing all sectors of ebook industry
… readers, publishers, SW, distributors, libraries, search engines, A11Y, etc.
… participation is also a success because all ey players are here
… (shows companies attending on screen)
… particpation is also international, 13 countries, 4 continents
… will allow to address int'l issues related to ebooks
… About highlights
… the event is divided in 5 sessions, 1 session per topic domain
… (lists all sessions, see agenda)
… Also a couple of keynotes
… we want this wkshp interactive so please participate
… we have then scheduled a lot of slots for discussions and feedback
… and also a wrap-up session
<gluejar> using #w3cebook on twitter
… "lead the ebook publishing to its full potential"'
… share info and expertise
… unlikely we'll solve all tech issues but at least we can discuss them
… create new work items specific to ebooks
… where and when to adress the tasks we will discuss: standard bodies or new WGs or or or
… Karen gave you wifi thingies, use IRC, the slides will be available on W3C site
… there will be bkfst tomorrow, break today is 4-4:30
… lunch not provided tomorrow by W3C
… all invited tomorrow evening to O'Reilly reception
… thanks to all sponsors of the event: Pearson, Adobe, Google, Microsoft and O'Reilly
… (my slides linked from agenda), thank you
Karen: introducing Jeff Jaffe, W3C CEO
… we have continued to grow under Jeff's leadership
<karl> s/Topic: Keynote - Extending W3C current work, collaboration with IDPF /Thierry's slides http://www.w3.org/2012/08/electronic-books/slides/S0-welcome/
jeff: Thanks Karen and Thierry
… let me be the 2nd to thank everyone
… in particular the host, collaborators and and and
… was asked to do 3 things
… since we're diverse community, first I'll provide an intro about w3c
… second, brief update on recent activities
… for you W3C members, that's not news
… but others need it
… third, some thoughts about why w3c thinks it is important for us
… w3c founded by Tim Berners-Lee, inventor of web
… still director of w3c
… simple and powerful mission, lead the web to full potential
… good job done but more to do
… member-based organization
… companies, research, academic, other standards bodies
… financial model comes from the largest companies called full members
… all the big names you know
… discussing with the whole ecosystem of Webthe
… professional staff to support and and the dialog with the industry, 70 on staff in 4 locations, latest host in China
… our major goal is to define tech standards
… tbl did not want companies to innovate alone and create wall gardens
… we then all work together
… contribute innovations and keep it open
… standards available to everybody
… 4 tech domains in W3C: interaction, Ubiquitous Web, Accessibility, Technology and society
… 45 WGs
… HTML 450 people in the WG, some other WG are much smaller
… liaison with 80 other orgs
… royalty-free patent policy is our major cornerstone
… we ask members to make royalty-free commitments
… it's more an int'l forum
… we have relationship with de jure standards bodies too
… lately, we've been working on a collection of standards: open web platform
… many different techs
… but also css, webapps for apis, fonts, device APIs, etc.
… large contrast with what we had 15 years ago
… rich interactivity
… very exciting capabilities
… interesting number of different devices
… was difficult to browse the web from a phone
… at CES everyone had a HTML5 set-top box
… the web is now ubiquitous
… making it work everywhere is a feature of the open web platform
… core supporting role for the trends of the industry and society
… survey done over a year ago, how many developers using html5
… now 150% :-)
… how many browsers will html5 in 2016: 2.1 billion
… Gartner surveyed tech
… number 2 was mobile apps
<azaroth42> How do they know 2.1B? Ran out their 32 bits ;)
… only the beginning
… as we move to next step, impact on businesses is amazing
… we talk to many people in indsutries
… next-gen technology
… business impact : opportunity for dialogs and business change
… should not surprise us
<tzviya> opportunity for businesses to change
… looking back, the basic idea of the web transformed sharing of info and al
jeff: what about publishing and the web?
… not strangers
… we publish web sites
… the web is a really nice tool
… introduced a new form of publishing
… able to reach more people in some more open ways
… when publishers needed add'l tools orgs created enhancements of what we're doing
… the web has democratized publishing
… every person is now a publisher
… we reached conclusion we need a more complete robust dialog with publishing community
… if we have a better dialog, we can make things happen better
… let's be at the front-line of publishing and web technology
… I'll talk about it at TOC on Wednesday
… I'm going to listen and learn today and tomorrow
… and will be your spokesperson at TOC
… will just mention a few high-level thoughts
… 4 categories for better dialog
… first, styling on the web
… nothing compared to classical publishing
… we need to learn from that
… second, more publishing will leverage the web
jeff: third, distribution
… fourth, the web is consumed differently
… that will change everything
jeff: that's all I had to say
… at the end of day, we discuss transforming businesses
… remember this is just the 1st meeting of that kind
… lots more conversation
… thank you
karen: introducing Bill McCoy, CEO IDPF
bill: thank you Karen and Jeff
… thank you all very much
… a privilege to be here and make this happen
… w3c process for worshop is pretty selective
… so thanks all of you
… many of us here may realize a hashtag is a subchannel of IRC
… #w3cebook is the hashtag
bill: warm-up presentation
… I'll be little controversial
bill: a word about IDPF
… 2/3rds of you are IDPF members
… 350 members
… all parts of the value chain
… mission is to foster an open ecosystem for digital publishing
… develop epub format
… dozen years
… one piece only of the big picture
… epub widely adopted for ebooks
… many ebook retailers distribute epub to consumers
… interchange format too
… conversion from epub to local format
… also delivery format
bill: needs to extend html to the future
… 2 years ago, kickoff meeting of epub3 in NYC
… (digression about Random House lobby)
… charter for epub3
… two distinct paths
… xml schema
… or build on the web
… at that time html5 support was minority
… adopting all of html5 would mean browser was required
… consensus reached eventually
… even Norman Walsh agreed
… 3 years later, still dealing with consequences
… was a brain transplant
… still parts difficult to deal with
… the publishing industry does not exist ina vacuum
… one click away from the web
… we cannot reinvent the wheel
… building on the web was a no-brainer
… if you look at the last 20+ years, 3 ways to deliver digital contents
… download files or apps ?
… not a topic for the next two days
… all valid ways to distribute contents
… depends on what consumer need
… sometimes one, sometimes the other
… we reinventing what we call books and magazines
… no need to argue about files#apps#...
bill: browser interface is what users see
… we're at the very beginning
… consumers are spending more time in apps and less in files
… controversial perhaps
… but not aguing against the open web platform
… web technologies are becoming common place
… so we're on the verge of success
… reusing tools and components across all modes of creation and distribution
… a universal platform
… the web platform
… not done yet
… we're also near failure
… not surprising
… the big risk is fragmentation
… look at webapps, many systems
… fragmentation already there in webapps
… in publishing we have our own suspects
… all html5-based
… but non interoperable
… many will support epub3 but no every feature
… so browser wars still alive and well
… of course, all propose proprietary extensions
… (digression about Adobe stealing from Xerox :-) )
… extract all the value from open standards
… taking advantage of it
… bootstrap your own proprietary platform
… shame on us as an open community if we let that happen
… even worse is monopoly control by a single vendor
… let's face it, we're almost here
… what can we do about fragmentation?
… 100 specs are used by us !
… we don't even have an exact count of them
… the open web platform is perceived as unrelated blocks of legos (shown on screen)
… we need a better architecture
… no enough vision
… what specs comprise the platform? too many WGs ?
… some WGs seems competitors
… html5 did successfully kill Flash
… but the open platform is not able to catch up yet
… the browser market share is today better than it used to be because no monopoly
… interoperable standards are here
… lots of issues to work out still
… 2 things we need to do, really:
… first, collaborate much better
… we're the bazaar not the cathedral
… we cannot remain in isolation
… browser vendors must NOT be the only ones dealing with the open web platform
… minimizing religious disagreements about secondary details
… (digression about Monty Python)
… not pointing fingers at W3C, same thing about IPDF
bill: we have responsibility and we're not there yet
… second, eliminate assumption that browsers are only to display web contents
… the OWP has to make documents and apps first-class
… everything can be served from the cloud
… we don't really know what the future experience will be
… we can also take a position here
… we simply have to do it to create momentum
… let's raise the bar of the OWP
… the browser wars have shown
… I'd like to see excellence
… question if the OWP should bother with requirements from commercial platforms
… does it matter if OWP is adopted by publishing and digital platforms?
… still doubt if W3C membership has embraced the notion of publishing industry+web
… going to be a win-win and we need to realize it
… full adoption is going to be large across various industries
<davidwood> Can http://www.webplatform.org/ play a role? Documentation of both the vision and how the components should fit together seems a possible direction to discourage fragmentation.
… high-design content needed
bill: interactivity and rich media
… w3c does not speak of semantic web any more
<davidwood> BUT THEY SHOULD
… learning material for education?
… semantic structure
… files, apps, websites
… multiple channels, interop
… int'l, global languages
… writing modes in CSS
… publishers need to represent content on various shapes and sizes of screens
… a11y is a critical focus
… print disabled people must have access
… we need user-friendly tools
… democratizing the web cannot only for coders
… benefits of the focusing on OWP are both ways
… it's not about publishing to raise the web, but vice-versa too
… I'm optimistic we can succeed
… universal OWP for publishing and web sites
<davidwood> Structured semantics are legitimately important (IMO) for publishers because the alternatives include fragmented individual solutions, unstructured semantics like tags and disconnected semantics like microformats.
… future of the web is up to all of you
… let's make next two days count
Håkon Lie, Opera: multicolor is CSS module
... resizing this, you can reset width
... number of columns changes dynamically
... basically say my ideal wideth is 14m
... one line of code
... we have this nasty scroll bar
... to read full article, you have to scroll up and down
... not idea
... so we added overflow page
... if I were on a tablet, I would use my finger
... this is one idea
... the Romans changed the world of publishing
... putting things on scrolls
... another magic things that happens when I combine floats
... with multicolumns
... why image is on the right
... also span two paragraphs
... moving from 4-3 columns, something gives
... image moves to next page
... don't have to write in detail, it happens normally
... We want more pictures and flow things to bottom, top, corners
... make elements span across all columns
... you see where I am aiming with this layout
... I replace those cats
... replaced them with sheep
... We don't know what The Guardian designer would have done
... this is all being done dynamically in front of you
... See the byline, second column
... It was hard to do
... pushed things aside a bit
... That's newspaper example
... this can also be done for magazines
... we can combine this with what we have in CSS
... the rotation, text shadows...still page layout
... and only ten lines of code
... Another interesting example is dictionaries
... many have text and images intertwined
... As I resize this
... the images make up white areas
... this image of Cato
... white space
... so find a solution...such as all pages on top
... but that doesn't work
... still use float and multicolumn layout
... but also use snap
... so it moves to its natural position in the layout
... this is very useful, especially in scientific publications
... I end with an academic paper
... This is a boring looking document
... but it has an important byline
... has been impossible to do on the web
... as I reformat for different screens, it changes
... I want it to be on bottom and in page mode
... There we are
... let me summarize briefly
... In ten lines of code
... I can replicate 90 percent of publications
... clean HTML, responsive design
... number of columns changes from tablet to mobile
... this works out the box
... Think we should have run-arounds
... and synchronization with baselines
... don't have those
... and select independent columns and pages
Alan: Next speaker is Vlad Levantovsky, Monotype
Vlad: Monotype is a big organization
... I found it to be useful to start with newspapers similar to Hokum
... see type
... two and a half inches in size with eight different type faces
... doesn't really help to convey information
... Nothing really changes when you look at the web
... as much type
... Jeff mentioned in his keynote, everyone is a publisher
... which is scarey
... So I'm going to talk about professional publishers
... not just what we say, but how we say
... I'm sure you'd like 'Harvey Davidson' in swirley font
... [a few other examples]
... Type face gives your message trust and integrity
... you can make it whisper or screan
Vlad: all these tricks can be accomplished
... Do we have web fonts? yes
... Embedded fonts? yes
... life is good; are we done?
... Not really
... Jan Tschhichold said, "Everything that counts in typography is...."
... Screen typography is very different from paper
... why it's different
... Here is how I want my type to be seen on the screen page
... when glyph outlines convert into pixels, it is blurry
... apologize for highly visual nature
... Quality of rendoring is big factor
... that is not easy to control
... those tyings are type features
... when you look at print publications
... today, most publishers use ligatures, numbering styles
... tools that publishers get accustomed to using
... Everything from the printing press will look identical
... to their true, original design
... with onscreen typography
... publishers need to understand the underlying screen limitations
... something that works well in print will not work on screen
... Here is an example of incorrect type faces
[Mad Men example]
Vlad: At Monotype we have been going through
effort of converting popular book faces into ebook formats
... shows changes of width, proportion, limitations of print display
... We have been developing completely new breeds of type faces for onscreen and ebook display
... Malabar is one such example
... see this onscreen in the Nook reader from Barnes & Noble
... gives crisp design
... and is popular with ebook readers
... That concludes my presentation
... I wanted to outline the problems, mostly in the professional publishing world
... We'll see more of them
... We need to communicate
... Publishers need to understand limitations of screen typography
... and web needs to understand their needs
... so we can accomplish both
Jaejeung Kim, KAIST
<tmichel> Enrichment of eBook User Interfaces: A Skeuomorphic Approach, Jaejeung Kim (Kaist)
JK: I am a user experience researcher at
... presentation is enrichment of user interfaces
... Reading a book requires a good presentation
... content layouts, font alignment
... another perspective is well manipulation of pages
... which requires a good user interface
... how the content changes depending upon user input
... Novels are mainly composed of text
... we read line by line in sequential order
... this is formal reading
<karl> ebook prototype UI by KAIST http://www.youtube.com/watch?v=rVyBwz1-AiE
JK: Text books are composed of text,
graphics...but not always in a sequence
... jump to references
... Newspapers and magazines are composed of all sorts of content
... we read without order
... so it's informal or casual reading
... eBook content are more than just text
... they are evolving to more interactive
... this is what ePub is aiming for
... it requires a more dynamic way of navigating through the content
... We conducted research to search for answer
... I will show you our design approach
... focus was not just eye candy or photo realism
... but to functionally contribute to the users' reading experience
... to bring print reading functional to touch-screen device
... Let me introduce you to two missing features
... thumbing through pages
... and temporal bookmarking
JK: our prototype is not based on any web
... Thumbing through is a four edge
... highlight area is called the four edge
... to thumb through
... you are able to perceive overall structure and content in a few seconds
... cannot use search
... have to go through pages to find a picture without explicit data
... use of this thumbing through was high in formal documents
... four edge UI is rendered on side
... touch dragging outward
... flips pages very quickly, in a few seconds
... and user is freely able to turn pages in a book holding position
... additional role of this four edge gives a tactile cue of page location, amount of pages left...plays a huge role
... also used to tag for location; a sort of bookmark
... we applied this in the ebook in the four edge area
... another feature was temporal bookmarking
... frequently done in book reading tests
... make a comparison among pages
... or stay on current page and get content from going back/forth to other pages
... use dragging gesture
... on release either return or stay the remote page
... Give you a demo
... This is thumbing through; second is temporal bookmarking
... compare pages and instantly return
... This video has had more than 500K views on YouTube
... Issues and requirements from Web perspective
... If it goes to web
... it requires a layout of the interface
... and rendering of page stack behind the current view page
... this cannot replace
... the slider bar
... to jump from page to page
... requires an HTML5 cache control to load pages
... also an API for placing additional features on four edge area
... like bookmarking
... and also flexible division of separate content on same page
... I have my device and you are welcome to try it hands-on later on
Alan Stearns, Adobe: introduces himself
… "Web versus eBooks"
… relatively new to both domains
… been working on CSS standards last two years
… epub features in CSS
… what epub needed for adaptive layout
… will detail my impressions
… great thing to see epub3 based on html5
… and the OWP
… the web ecosystem is so much larger than the one we have for ebooks
… it's going thru the transition now
… more widespread authoring skills
… Web+EPUB is then a good shift
… but what I would like to see is the two techs working better together
stearns: where interests converge, we should find a single solution
… where EPUB leads, improve the Web based on it
… for instance the EPUB content document
… we should work on it in the CSS WG looking at those requirements
… where interests diverge, make the web extensible and use polyfills
… the IDPF should create those polyfills
… packaging is a convergence
… offline apps and documents share a lot of things
… would be an awful thing to see them diverge
… we should work at converging them into a single baseline
… the Content document for EPUB was leading the W3C standards and used prefixed CSS properties to do that
… some people complained about using prefixed properties from specs in progress in CSS but I think it was fine
… showing what they actually need
… once you have prefixed properties, you have to push ; IDPF has to push
… to make W3C specs move along the REC track
stearns: we should prioritize based on that
… CSS TExt are crucial to EPUB
… needed also for the OWP
… still a Working Draft until last week
stearns: CSS WG meeting last week about it
… I'm guilty about keeping things late
… Want to push EPUB things need
… CSS 3 Speech
… there epub properties about this
… in CSS WG this ended up at the end of list of priorities
… only the editor has been pushing this
… so we need more participation from IDPF about this
… it needs to go to LC and needs Test Suite
… this particular lags because of lack of interest
stearns: that's one point of collaboration we can do
… CSS specs needs test suites
… and love more generally
… IDPF could contribute producing them
… naive understanding is that IDPF has not done that testing
… there could be more
… if there are EPUB tests we can go to, are this or that property available in EPUB readers?
… each viewer is not required to support everything
stearns: so different capabilities across readers
… now, polyfills...
… there are some things that could be added on top of OWP to support EPUB features
… for instance in JS
… would help viewer development
stearns: adaptive layout for instance, Adobe had a large JS library
… not really the way you want to do a polyfill
… smaller minimal chunks per feature
… not a comprehensive, too large library
… not everything can be polyfilled
… things from håkon's demo for instance
… some things should be prioritized in the OWP or made so they can be polyfilled
stearns: the CSS WG appears to me to come up with 80% solutions that are not extensible
… hence dead ends
… and then JS is needed
… we should have extensibility points so an ecosystem like eBooks can build upon our stack
… paginated views for instance
… would be a terrible failure if we come up with different solutions
… a bit about my own specs, Regions, Exclusions & Shapes, Page Templates
stearns: all things about adaptive layout
… find a good isolated feature for each piece
… CSS features EPUB can build upon
… I would like to see more collaboration and feedback about this
… CSS Regions in particular has diverged from original intent, from what you see in EPUB
… so complaining a bit : we need more collaboration
… when I brought proposal to IDPF, I got silence and splitism
… that attitude needs to change
… there will always be changes in a standards process
… you just have to accept it and live with it
… I'm not going to take more time
… I want Q&A now
… I want to raise the OWP to ebook standards
Karl Dubost: how many people know about polyfills and caniuse in the room? raise hands !
people raise hands
<tmichel> Polyfill about 30 %
stearns: polyfill 1/2, caniuse a bit less
stearns explains what they are
stearns explains caniuse
Ambica Desaraju (CourseSmart): 3 questions
… could bindings be an example of polyfills ?
… epub lists bindings as fallbacks for widgets the browser does not support
… for example slideshow
… bindings pulled out at run time
stearns: you choose different JS ?
... then yes
ambica: future of html5 appcache?
plh: working on it
… the appcache mechanism is broken right now
… implemented but broken
… meeting last week in London to find a new proposal
… work being done
… trying to find a solution
ambica: are mobile devices considered too?
stearns: documents and apps are all the same thing ; example of PhoneGap and polyfill model
… whenever we build somthg into PhoneGap, we want that to become obsolete and polyfills the way to go
ambica: howcome, you listed ten lines of coce
… what about responsive design for mobile devices, CSS Media Queries for many devices
Håkon: there were no MQ in my examples
… only multicol
… when you hit the limit, you need the MQ but that's a last resort solution
… we have a range here
plh: Philippe Le Hégaret, W3C
… thanks for bringing testing
… had a meeting in SF about that
… reps from mobile and non-mobile industry
… all had a their own profile of OWP
… they realized they share 90%
… lots of common interest
… if we're going to look at epub's profile, lots of common interest too
… so contributing to testing to OWP
<tmichel> Kim Marriott, Monash Univerisity, Melbourne
Kim Marriott : CSS has come a long way
… one of the things we miss in CSS
… will it do everything for EPUB?
Håkon: yeah, people always ask me for boustrophedon
stearns: I have a longer list than Håkon
Håkon: I think we should not stop before Guthenberg's bible
<tmichel> Marky Gylling IDPF
Marcus Gylling, IDPF: we have few efforts on testing
… support grid
… similar to caniuse
… test suite on github
ACTION mgylling post URL here
mgylling: we should find ways to share testing platform
stearns: we just had meeting 2 weeks ago about that
… we can coordinate
<abole> This BISG EPUB 3 Support Grid is set to be updated later this month...then again in April 2013... http://www.bisg.org/what-we-do-16-152-epub-3-support-grid.php
… about TTWF
… will help to hear from IDPF and vice-versa
stearns: Tobbie Langel, Facebook and now W3C fellow is now your contact
Peter Krautzberger (Mathjax): polyfills are very interesting and we call for more collab
… what can we do with MathML since we see no interest from browser vendors
stearns: not familiar with MathML
… what is the alternative?
… images ?
peter: we produce both and SVG output
… we can do that only if you have mathml support
… webkit has not a single developer about it
… so how can you push a standards ?
stearns: mathml is more complex than the example I have : balancing text
… adobe developed a polyfill for it
<karen> Daniel Glazman: If I can comment
<karen> ...we have an example in the market
<karen> ...about a large difference between WebKit
<karen> ...on Asian languages
<karen> ...only way to improve mark-up is to change it all
<karen> Kazuyuki Ashimura, W3C: Thank you for your great presentations
<karen> ...I was interested in Alan's presentation
<karen> ...ebooks services and devices
<karen> ...Opera had speech
Glazman: just use whatever rendering engine is available, that's going to introduce a shift in the market and other engines will respond
<karen> ...and speech API is implemented
<karen> ...W3C is working on speech capability
<karen> ...for web apps
<scribe> scribenick: karen
UNKNOWN_SPEAKER: ePub3 includes SML
... wondering about what type of extensibility should be for eBooks?
... is there anything specified; other options?
Håkon: You want to find common solutions to
problems without pollyfils
... sometimes will fail when moving to other devices
... want to identify the core problems
... certainly speech and audio are high on the list
... we did implement this with IBM
... we found we had a crash bug in the code
... and people had not noticed
Kaz: Do you have suggestions for speech interface, Alan?
Alan: I don't have enough experience to
provide an informed opinion
... anyone else who has experience in the audience
Gerardo Capio, Benetech
GC: we use Chrome speech to text capability
... which requires an extension
... they have been working on working group
... to put forward a speech API
... expect they will implement a speech synthesis piece
... may be available in Chrome without extensions
... And seeing other browsers like Firefox using it; will try to show a demo tomorrow in my presentation
Markku Hakkinen: Another part of speech is haptics
Markku: we have been experiment with this
... help students be more accessible
... and more engaged
... We are testing this
... would like to see how this works
... other technologies coming out such as haptics in CSS
... any others?
Gerrard: I'll try to do a demo
Ivan Herman, W3C
IH: There is an underlying issue
... not going into technical details
... that comes up with all the discussions
... that every decision is make on whether another browser is using something
... W3C, like IDPF is a member organization
... we work the members we get; those who are there
... in that WG, if only the big browsers are present, then they will take the decision
... that is the way it works
... The only way to change that
... is to have people who represent the users of OWP
... like eBooks
... not sure how we work that out
... there is an underlying thing
... and I can imagine same for IDPF
... we need the active presence and participation of the publishers
... to help turn the direction
... might be a fight
... with the browser vendors
... What triggered me was something said about WebKit
... not having MathML
... WebKit is a place to put code in
... if there are communities that want MathML in WebKit, they need to add it
Murray Maloney: has been doing the work for years
Ivan: Only an example
... underlying issue is the same...be present please
... have to work together on how to accomplish this
... we have to find the mechanics for this
Alan: You can talk about being present in the room, or you can cast it as engagement [laughter]
Murray: some publishers want rich media
... in multiple languages
... without having their assets multiplied
... I had my catch-all answer
... We have to "engage"
... maybe there is more information to give
Alan: I do know that there is a fight for the soul of sub-titling going on in W3C
Alan: VTT vs TTML
... there may be a tipping point you can influence by joining now
Daniel Glazman: a point about collaboration
... CSS Working Group designs specs but we are not the users
... We had a big divergence
... sometimes big arguments with the web designers
... Brad @ is helping us do the right thing in CSS
... for example if the grammar is bad
... or not the best one for designers
... you should do that, too
... Your presence is absolutely needed
... We all forget that most ePub viewers are based on two rendering engines
... Another way to help is to contribute code
... higher a developer, it's not that expensive
... compared to the publishing industry
... higher a developer
... W3C develops software...like the Validator
... it took years, do it
Alan: MathML is in WebKit
... decision of Safari and Chrome whether to release it
Daniel Glazman: fork it
Neil Soiffer Design Science: I want to ask Vlad about fonts
Neil: what are we doing about making math
fonts beautiful for the web
... and make sure there are all the technical symbols and math symbols
Vlad: We are not doing enough
... we should do more
... One of the messages I wanted to get across
... is to ask people what they need so we can work on them
... So this question is important to us
... not sure I can answer right now
Neil Soiffer: MathML Drop by Chrome...was disservice to math and engineering communities
Neil: Safari put it in even with security
... issue is not evil and they hate math
... but more that nobody cared
... I believe as a community
... publishing community needs to step up and say this is important
... and make it a priority
... someone needs to take responsibility
... no one wants to step up and make sure it's there
... It's really a disservice
... hope anyone in this group can help with that
Håkon: A colleague has identified a subset
... and then attach a style sheet
... I have not tried it
Neil: I was in original MathML but it's not
... our script takes huges advantage of CSS
... maybe redo some of those
... not looking at take this small thing
... have to download all sorts of fonts
Murray Maloney: I have been aware of the math problem since the mid-90s
Murray: multiple math societies complainted
... Many people in community have recognized this problem and have put effort into it
... @ Did the work on his own
... no compensation
... he wants one language as the one true language
... as the only one that is not supported properly on the web
... any time he went to somebody to get this code activity
... Seems there is always somebody smarter than you
... they think about it and it never gets done
... MathML work is done
... publishers should feel confident in their ability to publish math but they cannot
... reminds me of problem with HTML5
... when HTML5 WG would not recognize things that were not out on the web
... so the fact that publishers were using something within their walled gardens
... that did not count
... in all the years I have worked in standards
... is that publishers never step up
... you need to get somebody in the room
... and you need to start putting more content out on the web
... so the people who develop these tools
... can see that publishers use the b and the itag
... people who work with you say it's the right thing to do
... but browsers don't help
... get a membership in W3C and start screaming bloody murder
... if you don't, it's going to be programmers who don't know anything about publishing
Liam Quin, W3C: I have to follow Murray
... a quick comment
... This discussion of how we change things
... that is what this workshop is for
... it's to figure out how to change things
... ask everyone to hold in your hearts
... to notice all these things that need to be changed and why
... and think about how we can make changes
... to some of the things we have heard
... getting MathML into Chrome; where are the test cases
... how do we get test cases to rec
... not one asnwer to everything
... more use cases, more examples
... Keep in mind please, how can we change the future [applause]
Alan: Last question
... before the break
Thierry: break for 30 minutes at back of
... I would like to thank Daniel and Karen for scribing
... please see me or Karen and volunteer for the next session
... Also, if you want to do demos
... we have round tables during the break
... so feel free to use them
... We will reconvene at 4:30pm
Daniel Glazman: (introduction of the good
and bad of ePub)
... I implemented epub2 and epub3.
... I'm not pointing fingers, just talking about issues *we* have to solve.
... (going through bgee requirements)
... epub3 specs are based on several dialects. There are too many to deal with for being able to implement it.
... we could decrease that number
... There are also important changes between epub2 and epub3
... Some drafts are considered as Recs, that's an issue. There were no unitary tests for epub.
... There are a lot of inconsistencies, unspecified parts, which need to be solved.
... All implementers, authors, tutorials writers understand the spec and its meaning.
... (normative references tables which are not normative.)
... Some of the documents can evolve a lot, and so the references will not be valid anymore.
... It makes it difficult to base your work on moving implementations.
... Some ebooks will become irrelevant in a few months.
... The very first thing you hit in an epub is a Manifest… but zip already contains a Manifest.
... Is it useful to know the relationships between the files?
... many things are already given from by the viewers engine and their api already.
... Too many TOCs
... We just need one.
... Metadata in epu3 were a nightmare to implement.
... the relationships in between the metadata are scary.
... id/idref are too complex. It should not happened. It's very bad for UI. People do not understand them.
... Heavy usage of namespaces make the documents bloated.
... 1000 pages are opened in a few seconds. It's not needed to have namespaces.
... URI management in epub is a proof of the devil
... complex management of property vocabularies.
... "We may remove the prefix in the future"
... but it's not specified what you should do once the prefix is removed.
... Compatibility between epub versions is a myth.
... content documents: no default rendering.
... epub3 refers to html5 which is still in work in progress.
... extra schema for html5 that editing tools don't use.
... is epub:trigger needed? There's an implementation cost with it.
... AltStyleTags meaningful link classes. No notifications to the CSS WG. It reduces the space of class.
... CSS profile based on WDs which are not stable.
... epub should be only a packaging format with Web standards only.
... Should use only html5, allow both serializations?
... Some decisions in the spec lead to bad UI requirements.
... We have in the W3C to listen the publishing platform, but we need to do a few things before HTML5 Rec, CSS regions, Archive API for zips, etc.
... BUT we need the participation of the publishing industry.
... Get rid of proprietary XML dialects, and Epub core model and tests, tests, tests, …
Soo choi: Soo
Choi, epub production department at Harpers Collins
... Reaching the same screen in the digital landscape.
... Retailers specific formats are coming back.
... "Enhanced ebook"
... (in house workflows for the print graph)
... Print files are converted to ebooks
... ISBN assignment are part of the issue. Every file format requires a unique format.
... It has been very difficult to support the different devices.
... sometimes the retailers are having their own features and they do not tell in advance what will they support.
... There should be an enforcement of epub3 by govs, or organizations.
... (Proposal for a limited subset of specs)
... We need a consistent and positive experience between the author and the reader
<karen> Dave Cramer, Hachette
Cramer: we want more robust ebooks.
... case study L.A. Noire
... some viewers render things very differently even the simpler thing
... some fonts are used very specifically for the mood of the book. Monospaced font was not possible to use on ibooks until we found a hack.
... We want users to have the option to see it as designed.
... We do not know which reading system our books is on.
... web developers use UA string.
... mediaqueries are useful if it was not crashing
... @support would be useful if it was implemented.
... For interactivity the spec is quite light.
... epub2 was defined for things working in theory but not in practice.
... epubpreflight for checking what is supported.
Kim Marriott: from Monash University
... standards should not be only static. We are designing for the future too.
... What are the opportunities for the future?
... we want to have interactive and dynamic contents.
... we want things to be immersive to be able to live the contents.
... we want to be able to customize the content.
... including collaborative and continuous authoring.
... Multiple devices and accessibility.
... (summary of the automatic document layout discussed at ACM on document engineering)
... There are things which are already available in CSS, but not everything.
... (mentioning things like pdftex, indesign, tex, vdp, etc.)
... accessible graphic, haptic feedback. We need content and a standard for it.
... When making standards, we need to think about the future.
glushko: Robert J. Glushko, Berkeley -
Bridging the Gap between ebook readers and browsers
... collaborative teaching with a multidisciplinary textbook
... We can imagine a networked discoverable books with transcluded content.
... It requires to think about books differently
... a book can not be only on the Web, they should be of the Web.
... We want books are native Web fabrics.
... We want to be able to have Web books which are Web things.
... Ebooks should be a first class Web citizen.
... It should be linkable.
... It should not be publisher centric.
... How do we store content is an issue. There should be a browser agnostic system, where we can identify things with URI.
... we want to be able to cache the content, manipulate it, modify it and have sync from the client to the server and so on.
... There should be a browser export/import format.
<karen> Murray Maloney
Murray Maloney: I was surprised by your
comment on id/idrefs
... first class links, How do we point things?
Daniel Glazman: we need linking mechanism
and counting mechanism.
... things right now are working inside one document, not across documents.
... How do we do it? And how do we make it happen right now?
Mia Amato, Skyhorse Publishing
mia: How do we plan to handle QA in between retailers and different devices?
soochoi: There are multiple rounds of QA involved on different devices. There is technical QA.
mia: editorial are involved?
nick @ Rufolo: we talk about advanced features, but even on the basic features there are a lot of issues.
...without the kindle, we are not making money. So we need to be compatible with them.
<karen> Nick @ Rufolo
Nick: How do we put pressure on them?
cramer: I'm very pessimistic about Amazon
... It's sad.
glushko: textbooks are not likely to go to kindle more than other things.
<karen> Frederick Hirsch, Nokia
mccoy (hirsch too): What is the roadmap
mccoy: It's why we are here today.
… The full power of the Web platform has to be in books. We do not know yet how to monetize it.
<karen> Robert Sanderson, Los Alamos National Labs
Robert: We heard about content cache, appcache
Robert: Could you share how annotations should be carried on in books.
glushko: My books are designed to evolve.
Publishers think in release once and for all.
... The system don't work this way. It's sad. We can't do continuous annotation.
UNKNOWN_SPEAKER: it's complicated thing, I don't have a good answer
Robert Sanderson: Daniel, can you weigh in on this
Robert: helping the reader or publisher or
somehwere in between
... to add content to the ebook
... and how that might work with an extended API
Daniel Glazman: content is not a problem
... content and distributing it is a problem
... all the repository owners of documents have applications
... they know how to distribute a new version of an app
... should know how to distribute a document
... it's not server side, it's on client side
... book, you need a dif
<fjh> my question was (a) given the concerns with ePub is deployment happening and is there a roadmap to fix issues
Glazou: this is something a bit @
... on annotations, likely need a linking mechanism between packages
<fjh> and (b) what is the deployment situation for education and textbooks (given need for interactivity etc)
Glazou: if you download one package, you
need a link
... whether it's free or not
... renderer gets data from two different channels
... that is likely the right way to do it
Neil, design science
Neil: No representatives from Apple or
... a challenge, where are the publishers
... and where are the implementers to do the standards
... I know this has been a problem
... this is a pretty critical situation
David Cramer: Apple has participated to some extent in the ePub work
David: expect they are under interesting
contraints from their management
... I have no idea what goes on
... Amazon has had no participation as far as I know
... feel that they don't need to talk to us
... 'If you build it they will come'
Karen: we had a non-response from Amazon outreach for this workshop
Daniel Glazman: we met last week in Tuscon
for CSS F2F
... and Apple was there
... but some of their own features we would like them to submit, we did not see
... Apple has a way of doing things that belongs to Apple
... and they don't have the right to speak at conferences
... unless they are allowed on the conference basis
... what can I say?
Markus: In terms of bashing companies not
... Apple has been one of the earliest implementers of ePub3
... they are certainly a good citizen in the ecosystem
... everything is relative
David: I would agree with that statement
Karl Dubost: We talked about a lot of issues for publishers
Karl: but also issues for readerse
... I am a big reader
... I can put an image, notes, annotation into my books
... I may put two books side by side
... these are things missing in the ebooks ecosytems
... cannot make notes between two books
... I'd like to have a wiki book approach
... and edit content inside the book
... plenty of things we cannot do right now
... where it's a failure to the print platform
RobertG: any book with a URI can do this
Glazou: if a book has no URI, not fragments
Robert Sanderson: in terms of annotations, there will be more talk about it tomorrow
<glazou> glazou: nevermind, my bad
Robert: we want to speed up the process.
ddd, Google: I just want to clarify something.
...It was not just a liaison. It was a recommendation of the css wg.
Daniel Glazman: no
ddd: Maybe there was misunderstanding, but
there was discussion made at the css wg.
... search for alternate stylesheets.
<karen> Peter Krautzberger
krautzberger: One of the key issues seems to be to get the reading systems out of the way
krautzberger: It should not be only implemented but also overrided.
cramer: Something it depends on the constraints that devs have seen in the wild.
Daniel Glazman: you can't really. The issue is that there are competitive advantage in between readers.
David cramer: It's a battle in between two paradigms.
Daniel Glazman: the vendor can infer many
things about your reading habits.
... these data are sold.
... all vendors want to keep readers to have access to these data.
tmichel: we reconvene tomorrow at 8:30