Digital Publishing Interest Group Teleconference

30 Jun 2014


See also: IRC log


Bill Kasdorf (Bill_Kasdorf), Brady Duga (duga), Charles LaPierre (clapierre1), Dave Cramer (dauwhe), David Stroup (david_stroup), Deborah Kaplan (dkaplan3), Frederick Hirsch (fjh), Ivan Herman (ivan), Julie Morris (julie), Karen Myers (karen), Laura Dawson (dawson), Laura Fowler, Liam Quin (Liam), Liza Daly, Luc Audrain (Luc), Markus Gylling (mgylling), Michael Miller, Paul Belfanti, Peter Krautzberger (pkra), Phil Madans (philm), Thierry Michel (tmichel), Timothy Cole (timCole), Vladimir Levantovsky (vlad)
Alan Stearns (astearns), Madi Solomon (madi)
Markus Gylling (mgylling)
Dave Cramer (dauwe)


<trackbot> Date: 30 June 2014

<clapierre1> Zakium 1.650.644 is me

<philm> Hi, Dave

<clapierre1> Zakium aabb is me

<pkra> pkra is +1.617.858.aahh

<scribe> scribenick: dauwhe

mgylling: objections to approving minutes?

clapierre: my name was mshouldbepelled in minutes

mgylling: one can edit minutes directly in IRC

Ivan: one word or two?

clapierre1: one

mgylling: Are minutes approved?
... yes.
... Topic: summer break

<fjh> some editing tips here, http://www.w3.org/2008/xmlsec/Group/Scribe-Instructions.html (members only link)

mgylling: should we have a summer break during July?
... it would not prevent individual task force work
... any reactions?

<clapierre1> thanks

mgylling: suggestion was to start again on 11 August

<AH_Miller> I've took my vacation last week.

Julie: July and August are equally busy in the US
... July is worse for me

mgylling: Liza, Ivan, what do you say?

<brady_duga> Sorry, I am having phone issues and on my commute to work, so I will have to drop and won’t be able to rejoin

<Luc> I'll be away beginning of August coming back after th 20th

mgylling: we could reconvene on Aug 11 as heartbeat check

Liam: some groups have editing review meetings--review docs on IRC

Ivan: during aug 11 heartbeat decide what to do

mgylling: we'll reconvene August 11

liza: a11y task force has been meeting
... would that be good topic for next week?

mgylling: Charles, do you think that would be a good topic?

charles: next Monday would be good for me.

Liza: let's do that.

<Karen> +1 Accessibility topic on 7 July

Ivan: herding stray phone numbers

<David_stroup> 585 David stroup

Markus: heavy sigh

mgylling: Peter, the idea is to get quality time to understand STEM use cases work
... and to get suggestions you need.
... I don't have a plan for this
... I'd like to learn about the thinking and wiki work you've done so far

pkra: I added a few more math-specific use cases
... things like graphing and diagrams
... not sure which direction work should go

<tm> sorry I had lost my ADSL connection :-((

pkra: progressive enhancements fed into STEM stuff
... user agent improvements

<ivan> Chemistry use case

<scribe> ... new standards like ChemML are in scope? Yes.

<ivan> Diagrams, Graphics Use cases

UNKNOWN_SPEAKER: is mathml enough for science?
... mathml may need work to support k-12

<Luc> Sorry cannot make it.

UNKNOWN_SPEAKER: there were emails

<Bill_Kasdorf> actually it was MathML 3.0 that added the K-12 support

UNKNOWN_SPEAKER: tcole brought up contacting digital library of mathematics folks
... George K. added some notes on a11y pointed towards describedby and other ARIA things
... that's what I've gathered so far
... my questions aabout scope
... how do web components work here?
... for chem there's ChemML that might be typical use case
... if platform doesn't already support these standards, could web components help?
... where is the balance to be found?
... mathml is a good example. Is it a problem or poster child for web standards?
... it's sucessfull on creation side
... it's the xml and html way of writing math
... but it's a huge problem because browsers haven't supported
... web components could solve some problems, but make it harder to publish these materisals

<liam> [I believe safari has mathml support, and there's some mediocre support in firefox]

UNKNOWN_SPEAKER: where to draw the line between web components and standards work
... can we do everything in SVG? Yes, but it doesn't solve problems of publishers.

mgylling: it's a valid question
... if we look at from use case perspective, you shouldn't have to worry
... the first thing we have to do is describe how publishers want to use the OWP
... whether we have a name is the next question
... you can pretend to be ignorant about the solution space, and describe the problelm space
... you have updated the toc in the use case directory?

pkra: yes

mgylling: it might sound like a dry, academic answer to separate things like that
...: web folks doen't know what content folks want to do
... mathml is a good example. Great expectations among publishers, nothing from web world
... if we say web components, we can answer every possible question
... lots of hype right now
... with a good amount of scripting and connectivity you can make things loook like they are real
... declarative vs scripted
... should publishing give up on declaritive and go all-in for scripting
... so we'd just draw math on canvas and give up on mathml
... key issue is declarative vs scripted

timCole: one nuance... based on mathml experience, why would publisher see advanatage in using some of these technologies

<Zakim> liam, you wanted to point out that web components don't really address the problem

timCole: people ask why can't I use TeX

Liam: three things
... just two
... one: web components would address the markup issue, but we already have mathml
... the difficult part is formatting
... so you either recreate mathjax
... there's nothing in CSS to do math formatting without a lot of work
... the other part is JATS
... which is used in STM publishing
... as opposed to STEM
... STM uses MathML extensively

<Bill_Kasdorf> and BITS is the book counterpart to JATS

clapierre1: on a11y we have to make sure we don
... lose descriptive presentation for math
... we can't forget that

Bill_Kasdorf: there's now BITS: book interchange tag suite
... Mathml is part of that
... declarative vs scripted. I'm on the declarative side
... here's something that's richly described, then scripting can do something with that
... I have a preference for making available a resource that many things can be done with

mgylling: I'm a big fan of declarative, as it helps with a11y, and the presentation can be negotiated with the user

<pbelfanti> must drop from call - regrets

mgylling: scripted stuff can have user negotiation, but is more like black box
... it is a tricky question
... we can love declarative
... but if the cost and struggle to get something to work that way is too high
... will the publishing community adopt?

Bill_Kasdorf: if it doesn't have support, it's not useful

Ivan: to comment on web components; I'm not married to them
... I didn't mean to use web components for mathml
... I was wondering about chemistry or electronics
... on declarative vs scripted
... in an ideal world the reason I was wondering about it was exactly that
... it gives the end user a way to extend the declarative part of html who's behaviour is encoded in JS
... once those components are defined, the author would use the markup in a declarative way
... and the component/script would do the presentation
... for example, with ChemML it's a way forward without brower support for ChemML
... i don't know if web components are rich enough for these kinds of uses
... but may be worth considering

mgylling: the setup you describe is not that different from MathML plus mathjax

pkra: I had understood it that way, Ivan. My question came from Markus inviting me to not care about the solution
... I do believe web components will be useful
... my question: how do we combine this?
... what do we want from interesteed parties?
... our use case is going to be to use this markup but not as a web standard, but as a web component standard
... how do we get the richest feedback?

mgylling: I think so

pkra: web platform offers everything with mathml svg canvas, you'd have a really powerful base
... it comes down to making it usable
... what do we do on the standards side of this
... do we pick winners, identify the standards with the biggest market
... and combine that with ideas that can work as web components

mgylling: the pub community wants answers. what should we focus on?
... big-time publishers still provide math as bitmap images
... not because they're ignorant but because it's the only thing that works everywhere
... this IG can provide answers: you should use X for this problem
... another thought: the fact that web components exist doesn't matter in most cases because publishers don't have the engineering staff
... that will feed into a preexisting web component taht would render such content

<Bill_Kasdorf> +1

mgylling: it doesn't remove the need for standards, just changes where things are standardized, as a web component thing or as a markup language

Ivan: not either/or
... from author/editor POV, it looks clear that a declarative standard is necessary.
... who would do that is separate question
... if I would do that today, the browsers will not just implement it by themselves
... we've learned the lesson of MathML
... I'd look at another way of implementing
... web components might be another way of implementing something
... so these things can be done in parallel
... my question is still (for example with ChemML)
... is this doable with web components?
... it may be technically not feasible
... I don't know if that's true or false
... it would be very valuable to find out the answer
... I don't just don't want to believe that web components would solve, in case they would't
... Peter, you may know this

<liam> [I note that today there's JavaScript to render CML in SVG]

Ivan: the reason why the browsers have not implemented mathml is that they don't feel there's a market
... it's not a priority
... would a concentrated documentation of all the needs of the publishing industry overall
... including scholarly publishing (and not just books)
... and bring in business figures to show how important this is
... would this help browsers change their mind?

mgylling: we should talk about who to reach out

Bill_Kasdorf: AAP implementation project, Mathml was the highest priority from a11y people
... on the strategy of mathml plus canvas, I believe that just deals with display issues
... and not interactive issues?

pkra: Yes

Bill_Kasdorf: another benefit of mathml is that it does both things

pkra: there's a bit of a gap here, there's two separate standards
... content vs display
... presentation is already pretty rich, a11y tools can use
... there aren't authoring tools in content mathml

Bill_Kasdorf: presentational is the highest priority

pkra: it's easy to convert back and forth
... then you can talk about more advanced computation/display

<Zakim> liam, you wanted to discuss some people to whom to reach out to - CML: peter Murray-Rust (of course); STM publishing: http://www.stm-assoc.org/ and others; mathml semantic tools

pkra: I've given up trying to understand vendors
... lots of interest on browser implemtation side, just not enough to get the ball rolling
... I agree with ivan, making the business case would help
... and that's not just for mathml

mgylling: that should be the ambition of the use cases
... let's just describe what publishers want to do

<liam> [some people to whom to reach out - CML: peter Murray-Rust (of course); STM publishing: http://www.stm-assoc.org/ and others; mathml semantic tools will happen, e.g. from maple, wulfram, if needed ]

Liam: who to reach out to?
... someone mentioned CML, Peter Murray-Rust; there's an association of STM publishers
... Maple and Wolfram are market leaders for software

<pkra> add sagemath, iPython.

<clapierre1> Happy Canada Day (tomorrow July 1) :)

Liam: if publishing world said we need this, there are people with authoring tools

<liam> [merci]

Bill_Kasdorf: Two more... I can get people from SSP folks

<AH_Miller> How about Design Science and the work being done by Frédéric Wang?

Bill_Kasdorf: also Design Science

<liam> +1 Design Science

Liam: they also make word plugins

mgylling: the idea would be to grow use case collection

<AH_Miller> Frédéric Wang has been working on MathML for Gecko and WebKit.

mgylling: not only to convince browsers about mathml
... but to grow the whole world of scholarly publishing
... and we invite feedback from more people
... do you have things to add to existing use case collection?

pkra: I have some rough ideas but there's not a list
... let's reach out to everyone to get ideas
... two things: who's working on the task force?
... Tim Cole is interested
... especially on non-math stuff
... engineering has many higher needs

Bill_Kasdorf: I'd be happy to help, I have lots of contacts

timCole: I have lots of contacts, too.

Ivan: before we go, one more comment
... Madi and Bill did lots of interviews with people in the space
... which would be synthesized into a document
... something similar would be helpful here
... since we want a "business plan"
... backing that up with major players and companies gives it more weight
... so you might want to look at what bill and Madi did as an example

Bill_Kasdorf: it's all on the wiki now

mgylling: it's a good way to gather data
... it does take time, though

Ivan: Force 11 is looking at scholarly publishing

pkra: Summary: I'm in a good place right now; I have a lot go through
... can plan what you're describing, and come back with more use cases etc

mgylling: sounds good to me

<clapierre1> yes

mgylling: we're over time. Any final comments?

Summary of Action Items

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.138 (CVS log)
$Date: 2014-07-01 07:03:46 $