eBooks: Great Expectations for Web Standards, Workshop Day 2 - Tuesday 12 February

<abole> CSS Priorities

<abole> Regions,CSS Priorities: Regions, Exclusions, Page Templates, Grids, Flexbox, Writing Modes, Text, Fonts, Improved Page Layout

<abole> Publishers need to be engaged and stand up for themselves -- key discussion point from yesterday's session

<karl> Scribenick: abole

Ruby -- CSS Ruby is dropped (for now...bc no one is working on it at the moment...would be restarted if someone showed interest

That said, HTML Ruby IS being worked on

MathML -- some web browswers are not good at math....so what do we do about it? Testing and developers?

MathML -- consider contributions to the webkit

HTML -- adding extentions to HTML...next will have HTML templates...bad news: will have features that are added to HTML that aren't necessarily added to XHTML syntax

On AppCache -- see Web Application Working Group ...and possibly the TAG as well

Interoperability -- still facing interoperability problems -- how do you produce interoperable content for ebooks -- lots of EPUB 3 functionality that isn't supported on the readers -- need testing??

Interoperability -- Increase test coverage (clear need for this) -- improve test tools (framework and management; W3C does not have the proper testing tools) -- better documentation (if we want people to participate, we need to give them the right documentation)

Mobile & TV industry are main drivers of the need for testing

If you want to have an impact on the web you have to stand up for your use cases -- don't wait for others to address them -- publishers need to better engage the web community (and participate! again, strong theme from day 1)

Case Study: Web & TV -- produced set of use cases and requirements which are being addressed in HTML and device API

Case Study: Web & TV -- good example to follow?

<fjh> Device API (DAP) network service discovery, http://www.w3.org/TR/2012/WD-discovery-api-20121004/

Floor open for Q&A

Murray: Did you say there will be features put into HTML that will cause HTML 5 and XHTML to fork?

Plh: Yes, ... [but confused about the response]

Murray: We can put elements into an HTML 5 doc different from an XHTML doc that makes it so HTML 5 will have non "well-formed" docs? Is this a proper theme for this workshop?

Daniel: All of this has an impact -- ebooks should collect all formats -- no difference should be made between the formats

W3C should make no difference between the various formats that the authors can provide...

How can W3C influence standards develop? By addressing this HTML template issue head on....

<glazou> glazou: reminding you XBL in the mozilla world can attach behaviours and shadow DOM to both XML and HTML

Q: Another way to say it...has a fork been introduced? It IS still a question of whether or not.... Has this come up in the Working Group? What do they think of all this?

A: To a degree.

Marcus: IDPF is running a working group called Advanced Hybrid Layout with use cases around comics -- extenstions to EPUB 3 that will address issues relevant to comics ...these EPUB 3 extentions should enable connections with comics

Garth from Google: list of CSS priorities...this is an ideal list...arch of standards development at IDPF requires more developer participation...publisher participation has been heavy in the beginning with setting priorities, then reading systems vendors come in to do the development work...need more help from developers in all aspects

<karen> Tviya Siegman, Wiley

Publisher participation -- not all publishers have expertise with technical aspects of spec...they want to participate, just need the right angle, fit

<karen> Bill McCoy, IDPF

<karen> Daniel Glazman

<karen> Daniel: I missed your slide about CSS

<karen> ...it remains tremendously important point about problems with IDPF and W3C

<karen> ...we say yesterday that CSS must improve to support IDPF points

<karen> ...but is not the only user of the spec

<karen> ...61 documents in the reg for WG

<karen> ...and are roughly 20-25 people in the world

<karen> ...more than 2 specs per person

<karen> ...long specs requiring thousands of tests

<karen> ...Would like your take about the problem, what we could,should do

<karen> ...you keep pinging the co-chairs about what you are going to do to solve this

<karen> ...we never heard your point of view

<karen> PLH: recently the CSS WG

<karen> ...went through an exercise

<karen> ...not sure if they are adhering to these priorities

<karen> ...If we don't want to run the risk fo taking years and years

<karen> ...we have to find a way to follow those priorities and move things forward

<karen> ...maybe we have to drop things features to get done in a reasonable amount of time

<karen> ...Getting more people at the table could help, but also takes longer to reach consensus

<karen> daniel: Do you see W3C staff coming to the table

<karen> ...and help to prioritize

<karen> ...even if it's contrary to the strategic interest of the members?

<karen> ...Co-chairs are always fighting against needs of the members

<karen> ...commercial members are users and know what they want now

<karen> ...Vendors know what they will provide in the future

<karen> ...how are you going to make the prioritization

<karen> PLH: I think we need to have a conversation with our members and with the Advisory Board of the Consortium

<karen> ...I am not going to show up in the CSS WG and say what they need to do now

<karen> Alan Stearns, Adobe

<karen> Alan: One of things that holds us back is test suites

<karen> ...One of the things I have been working on is crowd sourcing

<karen> ...with test-the-web-forward

<karen> ...We have to take the prioritization exercise as input on what to focus the community on

<karen> Ivan Herman, W3C

<karen> Ivan: I would like to switch to slightly different topic

<karen> ...This remark is some thing I have heard

<karen> ...is that publishers don't have enough expertise to sit down with some of the web developers to have these technical discussions

<karen> ...It's very important to realise for our long-term goals

<karen> ...different groups have different roles

<karen> ...patent policy for example

<karen> ...If I look at the Business Groups and Community Groups

<karen> ...they get together people

<karen> ...to have, like for the TV industry

<karen> ...to get all the requirements together and then pass them along into the groups with the technial expertise

<karen> ...probably, and this is a discussion to have after this workshop

<karen> ...one way to go forward

<karen> ...is one of these groups should be established not only with IDPF but also with theother stakeholders to clearly formulate these issues

<karen> ...XHTML, HTML are some of the cases

<karen> ...Maybe some of the expertise will come forward

<karen> ...maybe after a year

<karen> ...there may be people involved enough to then join the WGs

<karen> ...I think there is a way forward

<karen> ...important to see there is a migration path to do that

<karen> ...we don't have to have people jump right away into the lion's den

<karen> Cindy Lewis, CarCulture

<karen> Cindy: I am in a CG

<karen> ...listening to some of the problems

<karen> ...my employment in reaching out to graduate programs at universities around the world

<karen> ...to help bring them on board to draft some of this

<karen> ...to enable people without technical skills to bring them in

<karen> ...would allow students to get resume credit

<karen> PLH: we need to wrap up

<karen> ...Ivan will do a final wrap up the end of the day

<karen> ...Thank you Alan and Markus

<karen> ...Our next speaker is Len Vlahos, BISG

<karl> scribenick: karen

Len: My talk is specifically non-technical
... and no slide deck
... my intent is to tell you about BISG and why it's relevant to you
... and to talk about the book industry
... and it's transition
... into the content industry
... The question from Wiley was really important to understand the need
... to relate developer communities to publishing communities
... BISG was first conceived in mid-seventies
... to conduct research
... with industry
... focus on supply chain issues
... and be nexus of industry thought and best practices
... we represent entire supply chain
... from authors guild...to retail distributors
... fully blended ecosystem
... print and digital
... purveyors of print fear we have left print behind and are digital obscessed
... some may think we don't understand benefits of digital and supply chain
... neither is true
... publishers sold more than $2b of digital content

<ivan> reference to Len's presentation

Len: retailers still sell to end-users
... but details are complex
... Three examples
... ePub
... Bill spoke about it
... important to think about it through the entire supply chain
... viewed as the industry standard
... our widely fractious supply chain came together around ePub
... it was a watershed moment
... Importance of industry acceptance to move things forward
... VHS and Betamax a good example [when it's not]
... Publishers can now see how ePub3 has been implemented
... keep it updated
... Success depends on more than just packaging and rendering
... ISBN published as ISO standard 2108
... to ID book products
... identifying books is straight forward
... some issues remain
... but are dealth with simply
... eBooks are something altogether different
... first ISBN assigned in 90s
... characteristics of ebooks are fluid
... may allow for printing or lending in one sales channel but not another, for example
... policy statement 1101
... document created with input from dozens of members ofindustry
... publisher may produce a single epub file
... business models in constant state of evolution
... committee pressure tested the tenants
... realities had changed since we first published
... ISBN have changed
... priorities have changed

<scribe> ...new revision will come up for review

UNKNOWN_SPEAKER: we reiterated and published; a collaborative process
... deliberations are both practical and philosophical
... what do we mean by @ product
... problems to product getting discovered
... identifiers is topic of DRM panel
... our industry view is DRM is transactional
... identifiers...need for industry consensus
... ISO is on five year cycle so comes up for review next year
... You'll hear from Todd Carpenter this afternoon
... helps keep US voice in that standard
... Final topic is metadata
... books face challenge of discoverability
... browsing experience has not been successful
... Core issues stopping it
... one is inaccurate metadata
... another is @
... direct correlation
... sounds obvious...see Nielsen study
... product metadata certification is needed
... before it reaches downstream partners
... we have a committee that looks at this
... all BISG committees are open to anyone who particpates in the ecosystem
... Angela Bole and I are happy to talk with you about how to get involved
... also involved is EDItEur who will speak later today
... participating in this conference today is important step
... and look forward to engaging W3C and this community
... in the future

<fjh> http://www.isbn.nielsenbook.co.uk/uploads/3971_Nielsen_Metadata_white_paper_A4(3).pdf

UNKNOWN_SPEAKER: all documents mentioned are referenced on web site
... distribution of content remains paramount
... whether print or digital
... we need to speak in one common language
... and continue our collaborations with other organizations
... Now like to introduce Janina Sajka

Next Session: Accessibility

Thierry: Before we start, I want to issue a call for scribes
... you will all benefit

<glazou> ScribeNick: glazou

<karen> ...we published a draft of yesterday's minutes

Yesterday's minutes available at http://www.w3.org/2013/02/11-ebooks-minutes.html

Accessibility panel, George Kerscher, Janina Sajka, Mark Hakkinen

moderator is Janina

Janina: good morning

… this session is about a11y

… topic of personal and professional importance to some of us

… colleagues of deep background here with me

… some of the best expertise on the planet here

… I'm chairing this panel today

… chairing several W3C WGs

… also background in publishing

… looking forward to reducing the gap

Janina: I think the last speaker was the Mark with an a11y perspective to attend a meeting
... George IDPF

… dedicated to success of digital publishing in general

… and more accessible content

… George will speak first

… me first

… Mark will end panel

George: story of a11y follows an interesting path

… "print disabled'" has evolved in our society : people who cannot read print

… the blind population is small, dyslexic larger

… but a lot of people can benefit from the features we introduce for these people

… first digital books 25 years ago and text-to-speech

… so we've been using digital books longer than anyone else

… a consortium was already serving that community to develop standards

… we turned to W3C for the standards in that domain

… Mark and I worked on SSML

… a11y guidelines

… IP policy

… the Daisy consortium has then been wortking closely with W3C

… sync'd text and audio

… as soon as we saw IDPF form, we needed to contribute and participate there

… work on all the specs that move forward

… when epub3 was being developed, we wanted to reuse W3C specs

… in other terms, don't reinvent the wheel, don't compete with existing web standards

… seemed to be a clear distinction between web people and publishers

… separation of presentation of semantics/content is absolutely needed

… footnotes don't even exist in html :-)

… EPUB invented then that part of the specification

… huge passion surrounding books

… magical belief that the digital book industry will be accessible out of the box

… so the community is interested in what'sgoing on

George: books in schools have to be accesible

… so people working on specs

… EPUB references html and we inherit from there

… the WCAG 1.0 apply

… 2.0 cover much more

… we needed to develop the EPUB a11y guidelines and we used the web ones to apply them to the publishing domain

… O'Reilly published this week EPUB3 best practices

… we have a check list for content producers

… available on idpf.org

… many features in EPUB spec support that accessibility

… very cool things with pronounciation lexikons

<karl> → http://idpf.org/accessibility/guidelines/ epub accessibility guidelines

… content guidelines for images

<abole> O'Reilly EPUB 3 Best Practices: http://shop.oreilly.com/product/0636920024897.do

… how to identify images descriptions

George: how to connect things, we don't have that yet

… a11y, Daisy Consortium has been a wonderful standards citizen

… mathml very important to a11y because people used images in the past

… now the math can be spoken

… navigation inside is possible

… so this is an overview of the things everyone will benefit from not only the disabled community

… how's the best way for people to use this

<Judy> Diagram Project that George mentioned: http://www.daisy.org/project/diagram

… data visualisation going forward is going to be great for people with disabilities

Janina: thanks george

… let me recap where we are

… what we have achieved and where we'll be going

… reusing standards

… is the bottomline principal

… a11y is a question of responsability

… not using standards costs more

<Judy> Janina's slides: http://www.w3.org/2013/Talks/02-12_ebooks_JS/

… and that means we need people in the entire chain to understand well what is required for a11y to succeed

… and we need the tools to make accessible content available

Janina: authors may have disabilities

… editors too

… we need to stay together to make this happen

… this can succeed

… benefits have been proven

… often, the accessible digital book is the only good option for people with disabilitied

… our standards MUST have a11y builtin

… DRMs should not break a11y

… the broadcast industry promised they will not break a11y, we'll check that !

… and of course, we need accessible content

… the distribution chain must be accessible

… allowing access to that content must also be accessible

… effective navigation does matter

… standards contribute to this if it's builtin so you can reuse that indutrial model

… all comes down to two things

we're not done yet

Janina: we're still working on images as George said

… issues with longdesc

… what happened with longdesc is quite similar to the mathml discussion yesterday

… people discussing did not see our interest because of low usage numbers

… remember this is a workaround

Janina: native SVG a11y

… engineering expertise joined SVG to add it

… we wanted to do this

… but lack of resources in the past

… SVG is a collection of lines

… and these lines form images

… where's the semantic

… it has to be specified

… or we have only a collection of lines and not a portrait

… a map of NYC for instance

… lots of geotagging

… many ways to represent data but how do you provide navigational information here

… need to develop the ability to associate a11y information to sets of data

… in ebooks you need a smooth and easy integration

… with ebooks we have new visibilities because of video in the content

… this is only me speaking for myself now

… if you have cookbook, you may want to compare sections

… or have two contents side-by-side

… asking content to reorganize itself on the fly

… let me briefly mention DRM

… lots of potential trouble there

… we could spend a lot of time on it

… having a11y friendly DRM is important

… there's going to be a balance between all actors here

… we need to avoid weighing on the people needing accessibility

… authorize things ONCE

… leave UI alone

… the national library service at library of congress is using standards-based systems

… content is authorized once

… (other examples)

… IndieUI (Independant User Interfaces)

… is a new WG

… part of a11y initiative at W3C

… extremely excited about the work done there

… inviting people there

… 2 deliverables

… Advanced deliverable and User content deliverable

… (first one) events and interfaces

… separate devices and multiple UI modalities

… "scroll-down" event should be disconnected from the way the event was indicated

… lastly the user needs modules

… all users can benefit from a11y techniques depending on user's environment for instance

Janina: (listing W3C resources, see slides)

… please read and we'd love to have feedback

<Judy_clone> http://www.w3.org/2013/Talks/02-12_ebooks_JS/#%2816%29

… we have a TF on html5 a11y

Markku Hakkinen now

Mark: testing

… Educational Testing Service

… embedded assessments within ebooks are a challenge

… there are some basics I'd like to focus on

… eTextbooks are here to stay

scribe: as a platform for elearning materials

… EPUB + OWP push to next level

… appearance of more interactive assesments

… for quizzes for instance

… the content itself can adapt based on results

… (shows a scan of a physics textbook)

Judy_clone, was mentioned Mark on his own slides

Mark: we need items to look the same across devices

… iBooks have a nice set of quizzes

… we cannot compromise a11y of contents to students

… legal requirements too here

… we need to see that when we're authoring contents

… the authoring tool must support a11y and we need support from readers

… WCAG must be supported too

… UAAG

… significant work needs on graphics and friends

… (shows examples of polyhedra)

… (with angles, vectors, segments...)

… (shows embossed print of a line graph)

… fairly expensive to print

… (shows a device with a large tactile display with a large grid, cost prohibitive)

… how can we use some of the capabiities in tablets today

… vibration feedback for instance

… move the finger and get feedback thru the vibration API

… haptic CSS

… authoring is an issue

… you only want to attach haptic styles to the data

… (example with haptic and audio feedback)

… refreshable braille

… connectivity between eTextbook and external world

… we may want to transform items on the fly

… how we return the results back

… in terms of standards, not only W3C and IDPF

… there are other standards orgs that matter (list on slide)

Markku: we're focusing on HTML5

… the place to go

… conformance to WCAG

… and other applicable standards

… keen on kbd interactivity and multimodal interaction

… kinect for instance

… compatibvility with assistive techs

… ensure assesment items are consistent in visual appearance across platforms

… and that's it

glazou: so webapps want native look & feel and a11y wants the contrary

Markku: we only want things to not change for a given platform

… it's more a question of familiarity for the students

… people need to understand what the content and the information is

last comment by George

Janina: the accessible representation may be different from the one you want to publish or the canonical one

… auhor proposes, users disposes

George: changing font for instance will change the layout

Neil Soffier: sometimes we need to have a difference preso

… MathML has an alt text tag and that will solve the problem

… but maths is 2D

… reading does not completely represent the formula

… 2x+1 is is (2x)+1 or 2(x+1) ?

… extra words to help the disabled community

… but oral noise to them

… more confusing than helping

… so don't use alt text tag

Janina: MathML in fact provides much of what we need

George: lot of work will be happening with MathML

… goes back to 20yrs

… present the information in the way the user wants it

… once you get expressions that are long, you can't deal with them in your mind

Janina: thank you all

<karen> scribenick: Karen

Session 4, DRM

Moderator is Jim Dovey, Kobo

Jim introduces panelists:

Youngwan, Samsung; Gerardo Capiel, Benetech; Oliver Brooks, Valobox

Jim: We're not here to come with a great new DRM scheme
... and not discuss the general issues and failures
... anyone can come up with a DRM scheme
... companies have their own forms of DRM
... as do publishers
... not a lot of benefit talking about yet another one of these
... everyone already knows enough
... I would like to talk about the technology underlying DRM
... encryption, signatures, identities
... we'll see more examples in a little while
... Underneath it all
... a DRM idea has four components
... User authentication, Device Authentication; Content Authentication, Action Authorization
... these boil down to authentication and identification
... Authentication is proving it's someone who they say they are
... I can say I'm Bill from CT; but you have to prove it
... to make somethuing happen
... gives you a guaranteed id
... any rights management system will require this
... to know who or what is something they claim to be is key
... and if they are the person is supposed to access
... or has paid to do what they want to do
... To read a physical book you can go to library for free or go to a book shop and pay for a book
... So if you're selling, you want to make sure your book is compsentated
... and make sure they are not breaking into it to alter it or make it free to others
... can be done without heavy encryption schemese

s/schemes

scribe: this scheme has worked well for Harry Potter books which are only watermarked
... also useful to idenfity a particular device for numerous reasons
... Perhaps a distributor wants to add special functions
... for example Kobo would encourage people to buy Kobo devices and have social media
... content you want to make sure it's not been modified
... important if including inside document encryption or watermarking information
... important to see if it's been stripped
... digital signatures tell if something has been modified
... we have X509 certificates to trace identity of the signature
... Authorization itself
... reading content
... pretty much has a demand
... having file is authorization enough, or having downloaded or having a decryption key to provide authorizatoin
... or have individual ways
... for re-downloading
... to fetch again content you may have
... Valobox speaker will talk about that
... Lending
... Libraries do lending
... I enjoy browsing books at friends homes
... nice way to discover new authors
... would be great to have that capability in the digital world
... Maybe certain libraries may want to limit devices
... much like DVDs
... certain eReaders known to be hacked
... Library could limit the type of device, especially if it's prone to free downloads
... One thing I'm hoping we can solve together
... many of these things require specific ePub
... nice to have a way to generate a single ePub
... and have libraries add a couple of requirements without breaking things
... If someone is not authorized
... content is plain text in a zipped file, you cannot stop, so you'll encrpt it
... and decrypt only after authorization
... Four things that might have identities
... Reader/consumer; purchaser; reading system or device and the content
... Look at Watermarking
... not lookingat who is reading right now
... not concerned about the reading system either
... content however
... don't know if watermarking schemes use digital signatures
... if we go to lightweight content protection
... useful to know who is looking at it now
... might be useful to distinguish who is reading or who is purchasing

Jim: I would expect encryption in ePub2 and 3
... encoding content
... and @ chains and more
... to validate the signatures
... private entities
... have as well, such as Adobe has its own rights schema and encryption tools
... Benetech has watermarking technology on their own
... IDPF is looking at this with a keen eye to standardize this
... so if you open an ePub the reader has some idea of what is going on
... and send an email to Adobe support for example
... rather than saying, "huh?"
... Thinking about an interoperable form of lightweight content protection
... XML=ENC and XML-DSig have been around for a whle
... believe they are candidate recs
... we have implementations
... Java enterprise

<fjh> Version 1.1 is in Proposed Recommendation

Jim: XML library from GNOME folks
... available in open source
... Great; seems like a good choice
... if you need to tweak it, the implementations can be complex
... I tried to find a new key mechanism
... not easiest thing I've ever done

<fjh> http://www.w3.org/News/2013#entry-9692

Jim: by a long shot..
... quite a lot of algorithms in these standards
... for small players, 20 algorithms
... might take a little while
... possibly someone could step in

<fjh> Functional explanation of changes in XML Encryption 1.1 http://www.w3.org/TR/2013/NOTE-xmlenc-core1-explain-20130124/

Jim: and make similar arguments to what IDPF did with SMIL in ePub3
... use these little bits of it
... reading systems are smaller target
... anyone making reading systems can handle authentication and rights definitations

<fjh> Functional explanation of changes in XML Signature 1.1 http://www.w3.org/TR/2012/NOTE-xmldsig-core1-explain-20121018/

Jim: usually this is kept proprietary
... not a great number of well known alternatives
... hopefully that is something we can collectively solve
... however, eReaders need to be cheap for people to buy them
... not going to sell 14m ebooks if they all require a $300 reader
... your market will be small
... people who want to read on Android will not come back due to slow reading
... ereaders built from best components
... If you look at implementing encryption and rights schemes
... signing and encrypting things on the fly
... books read on that system are not going to be easy to read
... important to think about this when defining standards
... none of these technologies says it should be used to restrict what people do
... Yesterday, Dave Cramer talked about how nice to have a device have these capabilities
... expand beyond X509
... is it an e@
... does it have a B&W screen; run Javascript
... we could provide this information
... When OCX files has manifestations of ePub
... we could have color sVG file
... or things designed for B&W screens that would render faster
... we can provide JS APIs
... and CSS thingys

[Daniel G smiles]

scribe: Provide hooks to content as well
... might enable publishers to think more about what they want users to do or not do with given books
... or help eleviate restrictions
... have to do it on their own
... Most important point
... Thik of the reader
... no one buys books because they have DRm
... if the reader is not sold
... you products are not getting sold either; don't forget that
... Now turn to Mingyung
... from Samsung

[Slides]

scribe: Like to talk about ePub3
... We have prototype about eBook WebKit performance

<fjh> Functional explanation of changes in XML Signature 1.1(corrected editors draft) http://www.w3.org/2008/xmlsec/Drafts/xmldsig-core1-explain/

scribe: Samsung has been delivering areas
... smartphones
... DRM is important to address
... user cannot read books when DRm is from service provider
... this is a problem to use ebook service
... need to discuss important DRM issues
... talking about performance issues
... We have found several performance issues
... difference between browsers and ebooks
... don't care about CSS or HTML5 or not
... we mostly tested ebooks with complicated contents with rich media
... to make pages for performance
... make it fast is challenging
... goal is to make all kind of pages ...
... process gaps...may show blank pages to users
... there are performance issues in CJK fonts
... layouts in those fonts take more time
... glyphs are bigger
... chapter navigation
... navigation through format makes scroll...large request for pages
... CSS straf exclusion

<scribe> ...new text alignment

UNKNOWN_SPEAKER: huge memory caused by heavy content
... as with encyclopedias
... all these contents should be loaded when ebook begings

s/begins

scribe: encoding in resources
... for rich media content
... power consumption
... magnified on publications...browser also has this problem
... We think web-based approaches could reduce market fragmentation
... eBook customers...hope social issues to be discussed and technical alignment with ePub and W3C

Jim introduces Gerardo Capiel from Benetech

GC: first point I want to make
... is that people with print disabilities use a wide range of technologies because they have a wide range of needs

slides shows a number of devices and software

scribe: let me show you a video

[video plays]

scribe: another technology I wanted to highlight
... we launched a new web reading experience
... a few significant things: It's an ePub reader
... and a precursor to webAPI
... This is bookshare
... will talk more about this
... before people had to download books, where we apply social DRM
... now they can read those books in their browser
... here is a physical science textbook
... to demonstrate the text-to-speech capability

<scribe> ...done by a volunteer

UNKNOWN_SPEAKER: this content is still protected
... a hidden finger print that ids user
... if I post it on the Internet, we could id who broke their contractual agreement with bookshare
... now that you understand some of technologies being used, let me go further
... The next point I want to make is that mainstream DRM hurts accessibility
... text needs to be readable by a wide range of technologies
... like test to speech or digital braille
... image description being read by text to speech engine did not come in original book.
... was written by a volunteer
... we need to create accessible alternatives
... for others to create
... for existing conent

s/content

scribe: images, long descriptions
... cannot be encapsulated
... pointers to alternatives to SVGs
... need MathML
... we have volunteers that take math equation images and scribe to MathML
... you might also need to add pronunciation into the content
... if file was fully encrypted, end user would be able to use
... We are largest library of accessible ebooks
... a wide range of offerings for different types of disabilities
... provide a wide variety of formats
... now accessible ePub3
... publishers also provide us ePub files
... so it's easier to work with content
... Most give us worldwide rights
... we have a large library for US and rest of world
... All enabled by the social DRM scheme agreed upon by all the publishers
... this seven-point DRM plan

[slide with 7 points]

scribe: benefits of social DRM
... clear accountability; meet our agreements
... users can use the AT that best meets their needs

AT=Assistive Technology

scribe: also lets parents and volunteers make it more usable
... provide some tools
... for example Poet, Toby from DAISY Consortium to make it more accessible
... volunteers have added more than 15K image descriptions to bookshare
... to wrap up
... My asks is that you consider social DRM as a viable alternative
... Look at some of work we are doing around adding accessibility metadata to Schema.org
... standards being used in ebooks
... you cannot find via Google easily
... you cannot find it
... when searching Google
... so by exposing info on accessibility through Schema.org
... we hope to make that reason less so
... Also encourage this community to educate yourself about speech API
... and alternatives to longdesk
... learn more about accessibility; take a look at the reading project
... thank you

Jim introduces Oliver Brooks, Valobox

Oliver: Provide a different take on getting around DRM and issues of it
... when you select a book
... even if a random purhcase
... if DRMed, you lock yourself in for a while
... Why is that a problem
... in digital book, unlike paper, the platform is the product
... when you lock yourself into a big, broad platform, you lock yourself it
... there are accessibility concerns and you may want to do more with your book
... How do we get around it
... idea if you buy something you should be able to synch to anywhere
... sych purchase information to A, C and D
... from user concern
... a Non-DRM book
... average reader doesn't know how to handle it
... if you are going to purchase something, are these guys going to support future devices?
... will they be around in a few years?
... So buy once, synch anywhere
... this is not trying to make a common dRM
... don't try to make individual files
... often dumbs down product
... What if we communicate the simplest item, the purchase
... ID the product and the user on each platform, so you can tell it
... does not have to be a common DRM, just a common identifier
... retail has sorted a lot of this out
... music, audio
... we don't have a database of every user, but they can identify themselves
... If I buy a book on "onebooks"
... then I buy on "twobooks"
... and connect by onebooks with my twobooks account
... may have seen user self identifying to synch
... then the two servers have permission to synch books
... books available on the two platforms now
... based on purchase info
... Not talking about ways to create metadata or files
... make something simple with connectivity
... for metadata there are many solutions
... What does this look like in practice; a server talking to a server
... pretty secure standard
... JSON is very simple
... share some ISBN numbers and HTTP
... Oauth used by everybody
... libraries can build in functionality; easy to implement
... data standard could be a simple hash
... with available products
... Here is the concept
... we're working with people like O'Reilly
... encourage you to look at newsletter and read more
... that's it

Jim: Thank you
... Does anyone have any questions?
... hopefully we have half hour for discussions

Erik Hellman: methods to enable library lending

Oliver: a buy one synch anywhere...very early days
... standard would be at its core
... just sharing access permissions
... easy to add vendor specific details such as a time limit for libraries

Frederick Hirsch: O'Reilly now lets you buy electronic books without DRM or sync

scribe: once you buy it, you own it
... not quite sure I understand whether they think the model is broken

Oliver: They are doing best they can
... when they provide to their customer base, programmers can manage themselves
... but average consumer does not want that hassle
... if something else comes out, you cannot guarantee that
... set up a simple synch
... don't need to do anything like always download or upload
... set up and it's tehre

s/there

Vlad Levantovsky: Question about finger printing

scribe: do you do it for all types of content
... or just images?

Gerard: we finger print all images

Vlad: Fonts?

Gerard: We do not provide font files in our packaging

Vlad: Aware of presentation of content

G: in future

Vlad: want to know what you do about it if they violate terms of the contract

Gerrard: We can follow up with member and see how that happened
... there have been a tiny set of situations for termination due to violation

Bill Rosenblatt: Giant Steps Media Strategies

scribe: are you aware of ultraviolet?

BillR: your scheme is pretty much ultravioletminus the scheme

Oliver: a bit more decentralized

BillR: yes, we should talk offline
... advantages to not decentralizing
... Ultraviolet moved to a more centralized model

Jeff Jaffe: question for Jim

scribe: phrase DRM is used by some in a narrow snese
... like how you talked more broadly
... wonder if it's time to talk about an even broader topic
... such as a web payments topics
... other things related to management for content
... mostly for Jim, but others can weigh in

Jim: this would be interesting
... but heard there are some compliance and legal issues
... I think it's worth having a conversation about
... to see if we can pick people's brains who know more about it than I do
... I think that could have merit
... anyone else?
... Any more questions?
... we have more time left

Liam Quin, W3c

Liam: I am interested in where you see relationship between DRM and lightweight
... approach like phone home for an upgrade

Jim: Two things don't have to be integrated
... one with another
... to cononicalize where something came from
... version finger print x; is there something I should know if there are new versions
... I believe we could bundle this sort of thing into a standard

Oliver: seems like something for the platform to perform as a common fingerprint

Taichi Kawabata, NTT:: I am interested in national libraries

scribe: how national libraries
... some will have to require legally publishers to put books into libraries
... will be kept for hundreds of years
... how can DRM survive such a long time

Jim: Good question
... this is the domain standards bodies
... make sure DRM runs as long as required is not governed entirely by proprietary formats
... so someone reading it in 40 years time
... Hope LOC might have their own ideas on how to best protect things
... would be interesting to know what sort of approaches are being taken

Oliver: hope library would have trust to have the non DRM file

Kawabatasan: Maybe over time
... thank you

Frederick Hirsch: Some people may have decided that plain text is the best way to handle it

scribe: when you go back to older files
... XML encryption

Jim: Maybe 2.0

Frederick: There is an algorithm with 1.1
... 2.0 has stopped due to lack of implementation
... it's NCR stage and may be a note

Jim: I should talk to you since I've implemented one of those

Frederick: that would be timely

Kazuyuki Ashimura, W3C

Ashimurasan: Web TV has discussed media extensions
... I think some input from ebooks world would be very useful

Jim: media extensions?

Ashimurasan: yes, basic way to handle in the web browsers

<karl> → https://dvcs.w3.org/hg/html-media/raw-file/tip/encrypted-media/encrypted-media.html Encrypted Media Extensions

Ashimurasan: as you mentioned ebooks not equal to browsers

<fjh> pointers to XML Security 2.0 work: http://www.w3.org/2008/xmlsec/wiki/PublicationStatus#XML_Security_2.0_publications

Ashimurasan: but please consider the technology

Gerrard: Those are areas I need to look at
... the reading project demonstrates that you can create a very good
... reading experience through the browser

<fjh> XML Security 2.0 is currently in CR but progression depends on implementation

Gerrard: there were some challenges, particularly with really large text books

<fjh> please let me know of any implementations

Gerrard: and older computers loading large books are a problem
... we support in ePub3

jim: Something I have thought about earlier...reading web project
... not sure if there is a standard JS way to deal with digital signatures
... maybe find an encryption API for browsers

Gerrard: there is some work being done in the human rights context
... encrypted web based chat
... I have seen some proposals out there around Javascript APIs

Ashimurasan: encryted media extension is not mechanism for DRm
... but just a basic primer to call it
... thank you

Philippe Le Hegaret, W3C

PLH: there is work within consortium on Web Crypto API

Frederick: IETF JOSE group is developing the equivalent of XML Signature and Encryption for JSON
... W3C Web Crypto group is working on underlying crypto technologies for Javascript

@: audio file is interesting

scribe: synchronizing
... must be id'd
... ID shoudl consider some encryption systems
... or encrypted HTML5

Gerrard: that would be a challenge
... not using media overlays in ePub3
... will look at that in the future

Jim: seems we are done
... thank you to all panelists and the hosts

<fjh> http://tools.ietf.org/wg/jose/

<karl> === LUNCH BREAK ===

13:30-15:00 Session 5: Metadata, Annotations

<karl> scribenick: abole

<karl> http://www.w3.org/2012/08/electronic-books/agenda.html#Session-5

<karl> « Annotation is considered to be a set of connected resources, typically including a body and a target, where the body is related to the » → what is annotation?

<ivan> slides on open annotation

<karl> * Open Annotation Community Group Position Statement, Robert Sanderson and Paolo Ciccarese (Co-Chairs of the W3C Open Annotation CG)

<karl> → http://openannotation.org/spec/core/ Open Annotation Core Model

<karl> * Bringing Magazines to the eBook Channel with EPUB 3 and PSV, Peter Meirs (Time Inc)

<karl> → http://www.w3.org/2012/08/electronic-books/slides/S5-Meirs.pdf

Nope. The room is cold. ....

<karl> * Developing and Applying Rights Expression using ODRL and RightsML, Stuart Myles (Associated Press)

<karl> → http://www.w3.org/2012/08/electronic-books/slides/S5-iptcsemwebstuartmyles.pdf

<karl> * Designing a Roadmap to a New Bibliographic Information Ecosystem: Where ebook metadata fits in, Todd Carpenter (NISO)

<karl> → http://www.w3.org/2012/08/electronic-books/slides/S5-todd.pdf

Todd Carpenter: http://www.w3.org/2012/08/electronic-books/submissions/webooks2013_submission_3

<karenm> Todd Carpenter: Open Community meeting in April in Baltimore/DC area

<karenm> David Wood: I believe everyone on panel could talk about metadata for the remainder of the day

<karenm> ...invite questions now

scribe: David Wood opens to audience Q&A

<karenm> scribenick: Angela

<abole> Karen -- will you be scribing?

<karenm> no

<karl> scribenick: abole

<karenm> just setting up for you :)

Eric Hellman, Gluejar...what's the best way for metadata to get from here to there?

scribe: with EPUB you can embed metadata
... legacy system is to send metadata to register and then have people pay

Mark Bide...answer is "it depends"

<erikmannens> ---

<erikmannens> On 14 Jan 2013, at 12:15, Jeroen Wijnen wrote:

<erikmannens> Hey Erik,

<erikmannens> Wij gingen eens samenzitten over onze engine he?

<erikmannens> Wanneer heb je tijd om ons te ontvangen :-)

<erikmannens> grtz

<erikmannens> Jeroen Wijnen

<erikmannens> Continuum Consulting NV

<erikmannens> Wetenschapspark 7

<erikmannens> 3590 Diepenbeek, Belgium

<erikmannens> T +32 (0)477 94 00 13

<erikmannens> jeroen.wijnen@continuum.be

<erikmannens> www.continuum.be

<erikmannens> ----

<erikmannens> Dit bericht (inclusief de bijlagen) kan vertrouwelijk zijn. Als u dit bericht ten onrechte hebt ontvangen, wordt u verzocht de afzender te informeren en het bericht te wissen. Het is niet toegestaan om dit bericht, geheel of gedeeltelijk, zonder toestemming te gebruiken of te verspreiden. Continuum Consulting NV sluit elke aansprakelijkheid uit wanneer informatie in deze e-mail niet correct, onvolledig of het niet tijdig overkomt, evenals indien er

scribe: you can embed all kinds of info in an EPUB, but not all...not price, for example, because it changes

<erikmannens> schade ontstaat ten gevolge van deze e-mail. Continuum Consulting NV garandeert niet dat het bericht vrij kan zijn van onderschepping of manipulatie daarvan door derden of computerprogramma’s die worden gebruikt voor elektronische berichten en het overbrengen van virussen.

Todd Carpenter...retailers have a sales model associated with metadata

David Wood...Americans tend to make things black and white ...no grey areas

Where should metatdata reside is a false dicotomy... a fuzzy boundry is the only way forward

Ivan Herman, w3c

scribe: where metadata recides is secondary
... horrified by amount of overlapping vocabularies...no idea which one to choose, under what guidence
... vocabulary world is a complicated one; this one seems to be one of the most complicated
... what should we do about this?
... Todd can't hope to solve the issue with NISO's Metadata Initative
... So, how can the different vocabularies be reconciled?

Mark Bide, EDItEUR

scribe: it's worse than that
... consider moving outside of text into adjacent media with more vocabularies
... extrodinarily hard to change the symatics used in a particular sector

David Wood, 3 Round Stones

scribe: it's better than that... there have been some successes
... over past 10 years we've created an international standarded, tied to the web, enabling us to create any vocabularies we want
... we have the basis for approaching the question

Todd Carpenter, NISO

scribe: we're too eager to address all the problems... this isn't how systems are built... it takes decades
... would prefer relying on LOC's catelogue of subject headings than "John on the corner"

Ivan Herman

scribe: one more comment

Nope!

Bob, U of California

scribe: always about how to protect publisher rights, as opposed to protect user rights
... how do you protect user rights?

Stuart Myles

scribe: it's possible to express this ... if people want to

Tom, Gent University

scribe: question about the problems of producing an ebook... any plans to adopt "this standard" in the ebook world
... "this standard" is the open annotations standard?

David Wood

scribe: working with provience data and ebook content and web content and migrating from internal providence vocabulary to the "prog" (???)
... suggestion that they submit an implementation report... will do

Bill Rosenblatt

scribe: Quesiton for Stuart
... AP implimented H-News. How does this relate?

Struart Myles, AP

scribe: no direct link between H-News and R-News and RightsML work
... both H-News and R-News do let you link using HTML links, but they're not the same standard

,,. they don't have the same use cases

Bill Rosenblatt

scribe: I heard h-news was to be used with webcrawlers

Stuart Myloes

scribe: yes, it can be used for this... so can r-news

Bill Rosenblatt

Studart

scribe: what other rights languages did you look at?
... Plus, CC-Raw (???) .... can provide more detail offline

Ivan Herman, W3C

scribe: a few years ago, W3C faced a similar issue on video and images
... lots of vocabularies -- a group was set up to look at the different vocabularies and provide a mapping
... not the goal to say that one of the vocabularies is "out" and no one should use it... put to simply provide a mapping so users can choose between the different vocabularies
... something like this for book metadata would be very interesting and important

Mark Bide, EDItEUR

scribe: been a lot of mapping (between ONIX and MARC, for instance)
... mappings are rarely one-to-one
... if linked content coalition, this mapping is its primary purpose so some mapping should come out of this

Ivan Herman

scribe: W3C would like to contribute... Done.

Peter (???)

scribe: [missed the question]

Peter

scribe: on authoring side, what do people prefer for authoring mathmatical content

Robert Sanderson

scribe: doesn't know... suggests Tholatic Krautzberger

Decisions on Future Work

<karl> scribenick: karl

liam: I will be highlight some of the things which have been discussed today.
... "So what?" What are we going to do next?
... After a few slides, we will have a discussion on concrete next steps.
... A few random comments:
... * Collaboration between W3C and IDPF
... * publishers can have a direct voice by participating
... * Stakeholders and user groups such as bookshare and accessibility
... Everyone can really participate. Get involved!
... Specific areas of possible work
... * DRM accessibility
... the barrier accross the door. We have to make sure it is NOT a barrier.
... There are things which are not known by category of engineers
... we have to work together
... * Extending WCAG and/or UAAG, be by extending or/and advocacy.
... Authoring guidelines, maths, graphics, navigation, wayfinding, etc. There are many ways of accessing the ebook.

* DRM and Human/ Digital Rights Management

scribe: In Canada, it is legal to copy stuff. Things vary from country to country and we have to cope with that at a large scale.
... We are not there at a point where accessibility is not an issue anymore. There is still a lot work to do.
... Encryption of the content, do we need a single system?
... Fingerprinting, social DRM. Google Images can help you search your images on the Web. Fingerprinting is a way to do that without having to rely on only 1 organization.
... Where to micropayments fit in? (millipayments)
... How do we handle long term archiving?
... Archiving a digital content for 50 years and then the readers are not working anymore.
... Maybe libraries should have access to unlocked content.
... Why an ebook is any different from a Website on a USB stick?

* Metadata

scribe: How to identify a digital object? (cf fingerprints.)
... annotations are feasible, but so far you can't highlight in a portable way.
... Relationships between PRISM, JDF, ODRL+RightML and others, how does it translate on the Web?
... Self-cataloguing ebooks would it be a solution?
... Matching ontologies, catalogs is one of the most difficult problem we have always in science, and computing.
... We are facing the same issues.
... User rights finally.
... You can have a voice. The future begins:

Question and Answers

Frederik: Is there a sense of prioritization?
... or a gap analysis?

liam: I do not believe the analysis of the various open web platform specifications or missing ones has been done. Because this is the place we should start at.

Judy: I'm not sure it is the best capture about accessibility requirements.
... wcag 2.0 is pretty baked and stable. We might do 3.0 in the future.
... But we could write techniques about epub
... There are areas there is a need for cross-fertilization work
... on IndieUI 2nd module, DRM, etc.
... There is critical work to do on CSS.
... such as haptic CSS and audio CSS.
... dynamic data-visualization could be helped with CSS audio.

glazou: do you mean aural CSS?

judy: yes.
... it was abandoned because it was not finished.

Markku: An education community group would be useful.
... We need to move further ahead than just simple webforms.
... We need science simulations, and have assistive technologies plugged to it.
... We want smart objects, smart images. We need to discuss in the space of publishing textbook which are accessible and richer.

liam: question to the audience
... how many people created an ebook?

(most people raising hands)

liam: how many people created an ebook reader?

(only 6 hands)

ccc: We need chemical formulae on the Web with maths expression.
... To the publishers in the room how many maths are in the top 3?

(3-4 hands)

ccc: maths not a priority

(3-4 hands)

ccc: Maths is an incredible expensive effort for some CATEGORY.
... You can get MathML out of wikipedia.
... (story of an accessibility issues with steps for accessing content)

liam: (story how captcha was broken by porn spam industry)

jeff: How can anything be more important than accessibility? How can anything be more important than css?
... We will have more priorities where the people get involved. Really involved.

Floor to Ivan Herman

ivan: We (w3c) didn't really know the publishing community. We did a good job by having the two communities discussing together.
... Thanks to the co-chairs.

(applause)

scribe: Thanks to O'Reilly
... Pearson, Microsoft, Google and Adobe (sponsors)
... What is the way forward
... Participation of the IDPF and BSIG and how they can participate to W3C
... but we didn't discuss a lot how W3C persons participate to these groups.
... Publishing industry should be present in the EXISTING groups. That's really the way to be heard by the current groups.
... There are a lot of groups.
... There's a strong need of cooperation for testing.
... We heard a lot of issues about interop and so on.
... Now that we know you, we will find you. (smile)
... Preliminary discussion on a roadmap.
... One of the most pressing discussions is the one with the CSS WG.
... We will try to organize a teleconference with the publishing industry and the CSS WG.
... To identify the most important issues.
... Daniel said "be rough, do not try to be politically correct."
... We might have to setup a Something Group. The group will look at the publishing industry as large.
... And what the W3C can do to better work with the publishing industry, what are the priorities?
... We should have more meetings in the future.
... We are playing with the idea, no commitments yet, there is a meeting in June in Tokyo.
... There will be a lot of people of W3C in Tokyo.
... Maybe we could have more participation from the asian community, with the help of IDPF
... There is a number of things going on at IDPF.
... Revisions of epub.
... There's a need of participation from W3C community in the work going on: Be W3C team, be some people from W3C community.
... Way forward beyond ebooks.
... Publishing is a lot more than ebooks.
... We will have a second workshop to focus a bit more on the backend side.
... In September 16-17, 2013, there will be two co-chairs, liam and Peter.
... We do not know yet where exactly, but Europe.
... And a separate workshop on news industry and magazine might be needed.
... Guardian, New-York Times, might happen in 2014.
... Possibly a workshop on metadata.

jeff: Setting up an informal discussion with IDPF, fine
... I don't think we can set up the Group ourselves (second point), the people in the room have to decide on what they want to do. Specifically member so fthe publishing industry

billmccoy: I do want to encourage if you are an IDPF member to join the conversation and to broaden the discussion.

kaz: regarding the meeting in Japan, we should consider about I18N in ebook services.

ivan: we will deal with it in time.

kaz: Just thinking about writing modes is not enough

ivan: we know where to find you ;)

liam: W3C could host a corpus of ebooks, sample.
... a few thousand of ebooks which are representative of what is done in the industry.
... It would be useful, even if you have to obfuscate some words by a bingo word.

ivan: There is a collection of sample ebooks maintained by ipdf

liam: different formats too.

marcus: +1 to what you are saying.
... I support your idea.
... the revision work is about to start, it is not yet started.
... It will be a minor revision.
... We are not into changing syntax so far, but we can start the discussion.

<karenm> Dave Cramer, Hachette Book Group

Dave Cramer: We almost need to annotate the spec to understand why it was written.

scribe: I would love to be able to track the discussions.

<tmichel> Karl: invited expert

<karenm> Bill McCoy: as far as background on ePub

<karenm> ...on Google site

<tmichel> gets in queue ...

<karenm> ...and an issue tracker, too

<karenm> ...Daniel pointed out a few errata also

<karenm> ...a history there

<karenm> ...and public archive of ePub WG

<karenm> Ivan; for the minutes

<fjh> http://code.google.com/p/epub-revision/

<karenm> Karl: ePub3

<karenm> ...looking for collaboration

<karenm> ...one thing that W3C does well

<karenm> ...is the tracking of the discussions

<karenm> ...and the actions

<karenm> ...another thing in the spec

<karenm> ...with XML

<karenm> ...spec, there is an annotated version which is even more useful

<karenm> ...than the spec itself

<karenm> Ivan: Are we exhausted

<karenm> Liam: A quick question

<karenm> ...If we do future work

<karenm> ...Who here would like to come to future meetings

<karenm> ...DRM: 7-8

How many people are willing to send people to work on CSS: 6-7

scribe: metatada: 15
... accessibility: 12
... annotations: 12
... html5: 3

<karenm> David Wood: What about MARC? :)

ivan: Thank you very much for the meeting

[Applause]

<karenm> Liberty Theater is at 236 West 42nd Street in New York City.

- DRAFT -

eBooks: Great Expectations for Web Standards, Workshop Day 2 - Tuesday 12 February - NY City

12 Feb 2013

Attendees

Contents

Accessibility panel, George Kerscher, Janina Sajka, Mark Hakkinen

Session 4, DRM

13:30-15:00 Session 5: Metadata, Annotations

Decisions on Future Work

Summary of Action Items

Scribe.perl diagnostic output