W3C

- DRAFT -

XHTML2 Working Group Teleconference

23 Apr 2009

Agenda

See also: IRC log

Attendees

Present
Steven, Roland, Markus, Gregory_Rosmaita, ShaneM, Alessio, Yves_S
Regrets
Rich_S
Chair
Roland_Merrick
Scribe
Gregory_Rosmaita

Contents


 

 

<Roland> trackbot, start telcon

<trackbot> Date - 23 April 2009

<Steven> can anyone hear me speaking?

no, steven

<scribe> Scribe: Gregory_Rosmaita

<scribe> ScribeNick: oedipus

Agenda Review

RM: each item will have 30 minutes devoted to it

SP: Yves will be joining us on ITS integration at 1400h UTC

<Steven> http://www.w3.org/MarkUp/xhtml2/meeting/2009-03-26#how_to_incorporate_ITS

RM: what do we need to cover with Yves

http://www.w3.org/MarkUp/xhtml2/wiki/2009-04-FtF-Agenda#ITS_Integration

SP: 3 options: inline similar to inline styling; put in head as meta info, or use XLink

MG: should as to use LINK element not XLink
... XLink one of many ways, so not a hard and fast requirement; do we want to do everything? has its own ruby module, translation stuff
... elements in legacy documents without changing them through external XSLT; use XPath to indicate were should be applied; referencing implementations - required to prove works

http://www.w3.org/International/its/itstagset/ImpReport

SP: is ITS out of CR?

SM: yes, been a rec for quite some time

SP: if referring to existing technologies, is ok

RM: what is minimum we can do?

SP: LINK element to ITS rules

RM: still doesn't limit exposure - still have to understand everything in ITS
... should we be incorporating required ITS features
... what is appropriate to us

http://www.w3.org/International/its/tests/

http://www.w3.org/TR/its/

MG: ok to just pick a sub-set

RM: yes

MG: ITS namespace has own ruby module - no link or dependency on Ruby spec itself
... should we consider switching to ITS 1.0?

SP: asked about this - also have own SPAN element

MG: correct

SP: requires a SPAN-like and Ruby-like element -- provided in cases where don't have them, but if available can use
... should check that interpretation with Yves

SM: their Ruby brought into line with Ruby Rec at some point

MG: couldn't find any differences - inline ruby element names and behaviors seem same as in Ruby spec

http://www.w3.org/International/its/techniques/its-techniques.html#integration-its-xhtmlmod

SM: fundamental question: to what extent can we sub-set this given our current functionality
... are we going to incorporate ITS as part of their namespace or ours?
... my vote is: do it in their namespace

SP: agree

GJR: agree

SP: if just point to set of rules rather than inline, only case that arises is its:translate attribute
... propose bringing its:translate into XHTML2

SM: as an attribute in their namespace

SP: yes
... what we are missing is "translate"

SM: inline ITS is a sticking point -

http://www.w3.org/MarkUp/xhtml2/wiki/2009-04-FtF-Agenda#ITS_Integration

SM: not bring in all elements, but bring in things like "translation" referenced to their namespace

SP: questions: 1) does spec allow us to use our native SPAN and RUBY and dir; 2) are you ok with only external ITS; can we use "ref" with LINK to accomplish this?? what would be the rel value
... how do they do that in XLink

<ShaneM> <link rel="itsRules" ... ?

<ShaneM> oooh. use CURIEs <link rel="its:rules"

SP: could suggest its:rules "rel" value

SM: would have to declare its prefix vocab mapping
... also ITS namespace has to be the same

SP: already an ITS namespace with value rules in it
... its:rules element

SM: correct

<Steven> http://www.w3.org/2005/11/its/

SM: could show them how to annotate as RDF

<ShaneM> There's no there-there right now

SP: no, this is more a namespace question, although there is a GRDDL version i thought i came across

SM: like idea of using CURIEs - use their vocabularies and have it just work

<Steven> <link>

<Steven> <link rel=�ts:rules" href="...."/>

SP: what is missing is version info

RM: LINK in HEAD and nowhere else
... allow its:translate on one of our elements

SM: vocabulary concatonated to name

MG: can take Qname, right?

SP: syntaxically similar to Qname

<Steven> a qname is a subset of curie

MG: MarkB - advised use of Qnames along with role values - is that right?

SM: interesting question
... can put anything want in quotation marks, but conforming RDFa parser not going to allow triples
... Qname doesn't map to IRI like CURIE

SP: ah
... P3P has reserved value for "rel" to do roughly the same thing

<Steven> rel="p3pv1"

<Steven> (note version info)

http://www.w3.org/International/its/itstagset/ImpReport

http://www.w3.org/International/its/itstagset/ImpReport

http://www.w3.org/International/its/tests/

http://www.w3.org/MarkUp/xhtml2/wiki/2009-04-FtF-Agenda#ITS_Integration

SM: from implementation report and conformance language, we conform by virtue of having a native SPAN attribute
... one must be available: SPAN, RUBY; if Ruby used, declare inline, if SPAN used, use as inline element
... did provide us with module as per our request
... also support ITS in M12n?

<markus> Birbeck discussing QNames in @role: http://internet-apps.blogspot.com/2008/05/something-which-was-part-of-early.html

<Roland> <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.1//EN" "http://www.w3.org/TR/xhtml11/DTD/xhtml11.dtd">

<Roland> <html xmlns="http://www.w3.org/1999/xhtml"

<Roland> xmlns:its="http://www.w3.org/2005/11/its"

<Roland> xml:lang="en">

<Roland> <head>

<Roland> <title>its</title>

<Roland> <link href="../its.." rel="its:rules" type="??" />

<Roland> </head>

<Roland> <body>

<Roland> <p its:translate="no">Grrrr</p>

<Roland> </body>

<Roland> </html>

proposed questions now at http://www.w3.org/MarkUp/xhtml2/wiki/2009-04-FtF-Agenda#ITS_Integration

GJR: apply to XHTML2 only or to 1.2?

RM: first agree on framework, then retrofit

shane, Yves will join at http://www.timeanddate.com/worldclock/meetingdetails.html?year=2009&month=04&day=23&hour=14&min=00&sec=0&p1=136&p2=179&p3=215&p4=248&p5=283

RM: other i18n issues to discuss?

[none logged]

Agenda Shaping

RM: order of agenda items: Access first?

<Steven> http://www.w3.org/MarkUp/2009/ED-xhtml-access-20090423/

SM: clear to move Access to second LC - put out new draft this morning and updated disposition of comments

http://www.w3.org/MarkUp/2009/ED-xhtml-access-20090423/

<Steven> http://www.w3.org/MarkUp/2009/xhtml-access-lc-doc-20090415.html

SM: happy to let draft go to LC

RM: just needs a bit of editorial tidying, but ready to go, so should start process

SM: yes
... chicken and egg problem: Access refers to CURIEs CR draft; changed CURIEs so doesn't have prefixing mechanism, putting onus on host language; Access isn't a host language, so we should be ok

RM: assuming HCG has no problem with this tomorrow, can immediately ask for XHTML Access to go back to Last Call

SM: should be more specific about whom we need to comment on Access

RM: have to address concerns brought up by PF and i18n

SP: some SVG stuff as well?

SM: no last call comment from SVG in disposition/issue tracker

GJR: PF has a long standing clarification request about abiility to mix targetid and targetrole in the @order attribute
... PF wants to know if is permissible

PFWG ACTION-211 - query XHTML2 and SVG on wheter mixing roles and ids with targetrole and targetid for @order is ok or problemmatic

http://www.w3.org/WAI/PF/Group/track/actions/211

<ShaneM> we got your voice mail

SM: in ACCESS element description must have targetrole OR targetid specified - if not present, mapping ignored

<Steven> sorry, problems with phoen here. Trying another

SM: not permitted in Access element - can have one or the other, can't have both

GJR: will communicate that to PF

<Steven> oh no

RM: user agent must only use targetid attribute values

<Roland> "If a targetid and a targetrole are both specified for an element, a user agent MUST only use the values from the the targetid attribute."

RM: have resolution from 2009-04-22 telecon to cycle Access Module back through Last Call

CURIEs

http://www.w3.org/MarkUp/Drafts/Overview.html#curie

http://www.w3.org/MarkUp/2009/ED-curie-20090422

<Steven> Gregory: When I go to banking services, the forms are half declarative, and half procedural

<Steven> ... I need a role value to express this

<Steven> ... some people are worried this would hold up ARIA

<Steven> Shane: Add it to the role module!

SM: nothing to do with ARIA - just define "form" in Role Module and be done with it

GJR: fine with me

RM: return to ARIA later;

SM: updated CURIE spec - removed xmlns stuff and produced new editor's draft

http://www.w3.org/MarkUp/2009/ED-curie-20090422

SM: changes CURIE into a datatype with a defines mapping with value space
... does it poorly, but does the job - can't find a better way anyway
... may get push-back from people who feel this change is too large
... already have 1 comment to that effect

MG: wasn't there an email that said up to host language to perform binding, so how can one have generic CURIE processor

SM: can't
... always permitted alternate prefix mapping mechanisms

SP: introduced to open hole to allow it to be inserted into HTML5?

SM: what?

SP: leaving it to host language

<Roland> http://lists.w3.org/Archives/Public/www-html-editor/2009AprJun/0009.html

SM: no, another objection; didn't want to use xmlns as prefix mechanism

MG: concerned about Role Module forcing HTML vocab to be default one
... ability to redefine vocab is separate but legitimate issue

SM: objection from a reviewer that didn't want to be forced to use xmlns - we thought that reasonable

SP: strong argument against having to go to LC again

<markus> http://lists.w3.org/Archives/Public/public-xhtml2/2009Mar/0081.html

<Roland> http://lists.w3.org/Archives/Public/www-html-editor/2009AprJun/0012.html

SP: if reviewer still objects - can state can have 1 or the other, but not both - going back to LC won't solve that problem

SM: generic processor never practical

<Roland> Includes: " I don't see how you can get out of CR to PR, looking at your implementation report. At this stage, I'm now asking Sean, my AC rep, to oppose such a transition."

SM: should finish collecting tests into a "test suite"
... do we need a "test suite"?
... for CURIEs? think we have one anyway -- just point to RDFa and be done with it
... could identify sub-set of relevant tests

RM: wasn't there a concern about a feature that RDFa doesn't use

SM: datatype, which isn't a feature; a lot of CURIE features not used by RDFa -- allow alternate prefix mapping mechanisms, etc.
... no default prefix
... if CURIE starts with a colon, don't define meaning for that in RDFa
... or a CURIE without a colon - in car, can't check
... MarkB had deemed those "reserved values"

MG: so what if CURIE does start with colon?

SP: prefix without colon - section 7 in RDFa "Module" defines how work with CURIEs with rules
... can leave out prefix according to CURIE document

SM: should say "for each of those, host languages may define a rule or mechanism to establish default prefix"

SP: "when prefix omitted from CURIE" doesn't cover the colon question
... if just put colon in front, use default prefix

SM: right

SP: got it

SM: agree that there are aspects of CURIE that RDFa does not use; isn't requirement that test suites be comprhensive, but all features of spec have to be used in 2 independent implementations
... we need to decide if satisfied criteria
... but that doesn't necessarily require a test suite

SP: have usage in RDFa, OWL

RM: aren't talking about implementations of those languages are we?

SP: can point to implementations of RDFa as implementation of CURIEs

SM: agree

SP: have a dozen of those

RM: markup languages include: OWL

SM: OWL WG said won't use CURIEs

RM: including OWL is a bit of a stretch then

SM: yes
... rolled their own rather than reuse CURIEs

<Roland> http://www.w3.org/TR/owl2-syntax/

RM: Last Call WD - 20 April 2009 - should send feedback

SM: part of SemWeb activity, which is origins of CURIEs so makes sense for us to eat our own dog food

RM: strange that didn't request feedback specifically from us since working on similar mechanism
... can we get comment from RDFa task force?

<scribe> ACTION: Shane - get RDFa task force to contribute to, sign onto or issue CURIE comment on Last Call draft of OWL2 [recorded in http://www.w3.org/2009/04/23-xhtml-minutes.html#action01]

<trackbot> Created ACTION-76 - - get RDFa task force to contribute to, sign onto or issue CURIE comment on Last Call draft of OWL2 [on Shane McCarron - due 2009-04-30].

http://www.w3.org/TR/2009/WD-owl2-overview-20090421/#Documentation_Roadmap

http://www.w3.org/2007/OWL/wiki/Changes_Since_December_2008

http://www.w3.org/2001/sw

http://www.w3.org/TR/2009/WD-owl2-overview-20090421/

http://www.w3.org/TR/2009/WD-owl2-primer-20090421/

http://www.w3.org/TR/2009/WD-owl2-new-features-20090421/

http://www.w3.org/TR/2009/WD-owl2-quick-reference-20090421/

RM: doesn't seem like a coordinated approach within SemWeb

SM: CURIEs useful in SPRQL
... way to view CURIEs is as datatype with some requirements on host language

SP: just like a URI

http://www.w3.org/2007/OWL

OWL1: http://www.w3.org/TR/2004/REC-owl-features-20040210/

RM: comment date is 12 May 2009 - need to be expeditious

<Roland> http://www.w3.org/TR/owl2-manchester-syntax/

<Roland> http://www.w3.org/TR/owl2-syntax/

<Roland> Comment By 12 May 2009

SP: if don't use CURIEs what do they use?

RM: their own solution

SM: which is VERY similar to CURIE, but with further restrictions
... called "abreviated IRIs"

from W3 front page blurb: "OWL 2 (a compatible extension of OWL 1) consists of 13 documents (7 technical, 4 instructional, and 2 group Notes)"

OWL comment action item: http://www.w3.org/MarkUp/tracker/actions/76

SM: is DAISY using CURIEs?

MG: yes

GJR: on list for ARIA 2.0 - support for CURIEs to point to external resources

RM: can mention RIF - RIF core and RIF collection/dialect
... other implementations?

GJR: ARIA use pending status of CURIE document

SM: Markus, have a pointer?

MG: not public info yet

SM: need something to point to - could be email from you to public-xhtml2@w3.org

RM: easiest to send email to public-xhtml2 with pointers to code in that mail

ACTION - Markus: send email to public-xhtml2 about DAISY's use of CURIEs to public-xhtml2

<trackbot> Sorry, couldn't find user - -

<scribe> ACTION: Markus - send email to public-xhtml2 about DAISY's use of CURIEs to public-xhtml2 [recorded in http://www.w3.org/2009/04/23-xhtml-minutes.html#action02]

<trackbot> Created ACTION-77 - - send email to public-xhtml2 about DAISY's use of CURIEs to public-xhtml2 [on Markus Gylling - due 2009-04-30].

RESOLUTION: request CURIE to be transitioned from CR to PR

trackbot, pointer

<trackbot> Sorry, oedipus, I don't understand 'trackbot, pointer'. Please refer to http://www.w3.org/2005/06/tracker/irc for help

[ADJOURN UNTIL 1345h UTC]

ARIA 2.0 CURIE Support Issue Tracker entry: http://www.w3.org/WAI/PF/Group/track/issues/285

<ShaneM> it occured to me that xhtml 1.2, xhtml 2, and xhtml modularization 2 also all use CURIEs

<ShaneM> oh - and XHTML Role and XHTML Access and RDFa of course...

http://www.w3.org/WAI/PF/Group/track/issues/285

http://www.w3.org/TR/wai-aria/#roles

SM: does aria's roles allow CURIEs?

"The applicable ARIA role is the concrete ARIA role whose name is matched by the first token in the sequence of tokens in the role attribute value which matches, on case-sensitive comparison, the name of any concrete ARIA role."

"A concrete ARIA role is any of the role types described above except abstract roles (roles for which isAbstract is true) and the roles imported from the Role Module, that is those that are introduced in sections 4.4.6 and 4.4.7"

http://www.w3.org/TR/wai-aria/#host_general

<alessio> http://www.w3.org/TR/wai-aria/#marquee

An implementing host language will provide an attribute with the following characteristics:

* The attribute name MUST be role;

* The attribute value MUST allow a space-separated sequence of whitespace-free substrings;

* The appearance of the name literal of any concrete ARIA role (see section 7.3.2) as one of these substrings MUST NOT in and of itself make the attribute value illegal in the host-language syntax; and

* If the name literal of a concrete ARIA role appears as one of the substrings in the space-separated list in the role attribute, the role MUST be processed in accordance with this specification.

SM: essentially CURIE-type mechanism by another name

ARIA 1.0 "Following the Namespaces Recommendation [XML-NAMES], the namespace name for these attributes has no value. The names of these attributes do not have a prefix offset by a colon; in the terms of namespaces they are unprefixed attribute names. The ECMAScript binding of the DOM interface getAttributeNS for example, treats an empty string ("") as representing this condition, so that both getAttribute("aria-busy") and getAttributeNS("", "aria-busy") acces

RM: request transition for CURIE, make LC comments to OWL2 to try and get consistency in SemWeb area

<Steven> Yves, we are ready when you are

<Steven> YvesS

<YvesS> ok, I'm getting there.

shane, http://www.w3.org/MarkUp/xhtml2/wiki/2009-04-FtF-Agenda#Issue_7659_and__Issue_7656

SM: should announce in rdfa.info that DAISY uses RDFa

MG: will go public with first draft before summer - nothing to look at at this point
... except for source code repository

ITS Integration Discussion with YvesS

yves, http://www.w3.org/MarkUp/xhtml2/wiki/2009-04-FtF-Agenda#ITS_Integration

background: the WG has identified at least 3 questions for Yves:

*

o does ITS allow XHTML to use its native SPAN element and Ruby module?

o is external ITS support alone sufficient?

o can we use LINK with "rel" in HEAD?

1. does ITS allow XHTML to use its native SPAN element and Ruby module?

2. is external ITS support alone sufficient?

3. can we use LINK with "rel" in HEAD?

Subtopic: Reuse of XHTML's Native SPAN element or ruby module?

SP: want to integrate ITS into XHTML2 - wanted to get info straight from source's mouth
... can we use native XHTML span to satisfy ITS?

Yves: correct

SM: what about Ruby module

Yves: if aligned with ITS

SP: ITS copied ours, so should be ok

Subtopic: LINK versus XLink

SP: can we use LINK with "rel" to link to ITS rules document

Yves: good question; as long as documented properly, shouldn't be an issue; don't see any conflict; processor just needs to know what to do with LINK

SM: have other groups identified value that want to use for "rel" attribute
... want to have "rel" attribute that states HREF pointing at is an ITS rule

Yves: XHTML2 WG should come up with value for "rel" attribute - we will follow your lead

SP: preference is to use a CURIE

<ShaneM> XHTML2 WG Gets to blaze the trail again!

http://www.w3.org/MarkUp/xhtml2/wiki/2009-04-FtF-Agenda#CURIE

SP: since rulename is its:rules therefore rel="its:rules"
... any objection to using CURIE?

Yves: don't forsee problem

SM: is HTML5 group going to accomodate ITS?

Yves: yes

SP: how?

Yves: explicit links

<ShaneM> HTML5 would not like use to use a colon

SP: if have to interoperate with their value, shouldn't be a CURIE, but keyword value

<ShaneM> rel="itsrules"

SM: if that is part of target audience, i agree -- if try to use CURIE, have vocabulary prefix issue to deal with, so prefer rel="its:rules"

<ShaneM> rel="itsRules" maybe

SM: without colon - then HTML5 can't complain about colon
... want maximum number of people to take advantage of and use ITS
... need consistent behavior model

Subtopic: ITS Rule Versioning

SP: do rules documents state version?

Yves: yes - describe way external tool can handle ITS document

SP: not worried that LINK wouldn't say which version pointing to?

Yves: no - customized tools is end goal

SP: don't want to have to change documents because of new version of ITS rules
... if allow LINK rel="itsRules" is that sufficient?

Yves: yes, provided can do the lookup

Subtopic: its:translate

SP: want to use its:translate, but already have own "dir" - can just use "translate" attribute

Yves: that is fine; map existing data categories/atttributes using ITS rules

SP: in its rules, can say attribute "dir" is same as its "dir"

<ShaneM> how about <span translate="whatever" dir="ltr">lala</span> ? instead of its:translate?

Yves: yes

SP: no problem importing its:translate - do you have preference if reuse yours or include a native XHTML translate attribute

Yves: use ITS

SP: not allergic to foreign namespaces

Yves: can map it

SP: thank you VERY much Yves for your attendance

Yves: HTML5 comment on question of value - suggest use "itsRules"
... will keep XHTML2 WG updated on negotiations with HTML WG

SM: create ITS module for XHTML2 spec that indicates bringing in itsRules attribute add to Common or Text module?

MG: inheritable, so that is reason to put in common

RM: text module

MG: 2 questions: ITS module and should it be in Common or Text Attribute collection
... where to put it is the ultimate question

SM: if add to existing module, i'd add to i18n module

[general agreement on wisdom of this]

SM: adds internationalization collection to the common collection

<Steven> http://www.w3.org/TR/xhtml2/mod-i18n.html#s_i18nmodule

SP: translate attribute marked as an issue in that very section

SM: for XHTML2, use i18n module
... architectual question: should i18n module always include ITS in M12n 2.0
... bring in rel="itsrules" and itsTranslate?
... i think the answer is yes

RM: don't see reason to say "no" at the moment

<Steven> itsrules

s/rel="itsRule"/rel="itsrules"/G

SP: why not put i18n and bi-directional text together in same module?

RM: and ruby??

SP: no, ruby has own attribute collection - referenced by indirection

SM: not a bad suggestion to collapse into i18n module, but don't want to consume f2f time on that editorial issue

RM: ITS use in LINK element with appropriate "rel" value
... investigate collation of ITS, ruby and bi-directionality into an expanded i18n module

Issues 7659 (AnneVK comments)

http://htmlwg.mn.aptest.com/cgi-bin/xhtml2-issues/XHTML-2.0?id=7659;user=guest;statetype=-1;upostype=-1;changetype=-1;restype=-1

SM: issued at when stopped last time - hypertext attribute selection

Subtopic: hypertext attribute module

SM: Anne asks why are needed at all

SP: hreflang doesn't solve problem - if multiple language versions of link and send dutch link to you who use russian, should get russian version
... hreflang not a new facilllity; always in HTML4
... problem - doesn't do anything - no better than comment
... if claim it is dutch than it better be "nl" - if don't want to do, then don't

GJR: hreflang is useful for accessibility because it prepares the Assitive techonology to make the appropriate natural language switch
... hreflang identifies natural language for natural language processing

SP: 2 problems: 1. not garunteed to be interoperable (will break if move URLs if web server doesn't use same method of specifying URLs
... 2. no way to garuntee that is possible - know at least 1 server where one can have several resources linked via URL but not to sub-resources

RM: content creator's choice whether or not to do that
... could make them available as choice or through content-negotiation

SP: hreflang doesn't break content negotiation
... hreflang offers extra functionality - if don't want to use, don't have to, but is useful functionality for some use cases
... example: if have on homepage links to different versions of homepage, this allows me to check if they are really there -- allows one to check links to ensure that what is supposed to be the dutch version is the dutch version

GJR: also good trigger to switch natural langauge for TTS, charsets for screen magnification, especially in absence of natural language declaration in document being referred to

RM: document benefits and extra functionality

<Steven> http://www.w3.org/TR/xhtml2/mod-hyperAttributes.html#s_hyperAttributesmodule

"This attribute specifies the primary language of the resource designated by href. At its most general, it is a comma-separated list of language ranges with optional accept parameters, as defined in section 14.4 of [RFC2616] as the field value of the Accept-Language request header."

"In its simplest case, this is just a language code, such as "nl", but it may also contain variant specifications such as "en-gb"."

RM: doesn't address why would use in concert with content negotiation

SP: current text isn't sufficient?

RM: don't think we cover value statement very well as currently written

example from document:

<p href="http://www.w3.org/2003/06/semantictour-pressrelease"

hreflang="fr">

The press release in French

</p>

RM: no complimentary documents that goes with XHTML2 that addresses how to use these things together
... will lose this if don't put into document

GJR: i have an action item to create a wiki page for Role BP; i could do the same for hreflang and content negotiation

RM: need "motivating examples"

SP: could put same example with <a hreflang="nl">Press Release in Dutch</a>

GJR: wouldn't you want <a hreflang="nl" xml:lang="nl">Dutch for "this page in dutch</a>

RM: content negotiation methods - match against what UAs do today

SP: if want to say this is version in french, so available to someone who wants french

RM: if all want to do is get what matches browser, don't have to worry; if want to give user option to choose language, then use hreflang

<scribe> ACTION: Gregory - create wiki page for hreflang and content-negotiation best practices and examples [recorded in http://www.w3.org/2009/04/23-xhtml-minutes.html#action03]

<trackbot> Created ACTION-78 - - create wiki page for hreflang and content-negotiation best practices and examples [on Gregory Rosmaita - due 2009-04-30].

GJR: make sure to address xml:lang in conjunction with hreflang and content negotiation

RM: done on hreflang for now?

SP: brief comment at end - target is CSS issue with pointer to old spec

SM: answer is "no it is not"

SP: right

SM: bi-directional text verbiage has typo i need to correct

http://www.w3.org/MarkUp/2009/xhtml-vocab-20090423.html

Subtopic: Edit

SM: Quote element doesn't have default presentation, whereas "edit" has one

RM: comment from DocBook about not having default presentation in quote

SP: reping CSS WG - ascertain if changed mind

RM: want to have consistency for content creators across languages

SP: can undo decision - is in current form due to comment and request

SM: have to revisit the issue anyway

Subtopic: Embedding Attribute Module and @type

SM: question is why need @type at all
... purpose of @type is so i can indicate to UA what types of resources are available that i want to use for my document
... UA should then determine intersection of what author wants and internal capabilities and serve the "best" one
... author in control of document and resources therein

RM: what about things embedded from elsewhere? don't necesarily have control over embedded content

SP: always been in HTML

RM: don't need old baggage

SM: some old baggage is useful and we have tightened usage
... @type allows me to serve image in SVG, PNG, JPG, and GIF - as author, want cascade: SVG, PNG, JPG or GIF depending on UA capabilities

SP: if don't want to use it, don't use it, full stop
... allows for more author control - increases author's power, doesn't decrease it

SM: what should http-accept header report

<alessio> provides a more stable object fallback

RM: don't have image module

<ShaneM> http://htmlwg.mn.aptest.com/htmlwg/xhtml-m12n-2/mod-csImgMap.html#s_csImgMapmodule

RM: first example uses NL - ok

SP: why arent' attributes applied directly to a list

SM: why do you need detail is comment
... if change to UL should solve problem

SP: <ul role="navigation"> right?

SM: sure
... don't call "LABEL" but "CAPTION"

MG: no LABEL element anymore

GJR: suggested use of LEGEND instead of CAPTION

SM: TITLE or CAPTION - different semantics

GJR: use old fieldset, legend. label model, freed from forms - irrelevant due to XForms module, so reuse LEGEND as generic labeller

Subtopic: Why Need OBJECT?

SM: something has to take PARAM

GJR: trying to get HTML5 to consider PARAM for EACH media-specific element

Subtopic: XHTML Scripting Module

SM: complaint that it was being dropped; we have decided to chop XML Events into event, handler and script

Subtopic: XForms

SM: don't understand "useable by everyone" forms instead

SP: send pointer to XForms in HTML tutorial

Issue-7656, Comments from Neil Katin

RM: some of these complaints/comments have been obsoleted by events

SM: most of them

http://htmlwg.mn.aptest.com/cgi-bin/xhtml2-issues/XHTML-2.0?id=7656;user=guest;statetype=-1;upostype=-1;changetype=-1;restype=-1

Subtopic: Example for hreftype

SM: will check to ensure that hreftype example is correct

Subtopic: Definition of Access

SM: editorial complaint

RM: shouldn't be problem - ripped out Access and made it its own module

SM: right

Subtopic: bidi text

comment: "(bidi text) there are two section definitions that are malformed: 15.1.3.15.1.1. Inheritance of text direction information and 15.1.4. 15.1.2. The effect of style sheets on bidirectionality"

SP: in HTML4 could only say deleted somewhere and inserted somewhere else -

GJR: proposed a for/id relationship for INS and DEL in Editing Module

http://www.w3.org/MarkUp/xhtml2/wiki/ProposedElements/ModificationMarkup

attributes: http://www.w3.org/MarkUp/xhtml2/wiki/ProposedElements/ModificationMarkup#Attributes_Necessary_for_Modification_Element.28s.29

GJR: lingering question: "is <DEL id="blah1">dig</DEL> <INS for="blah1">dog</INS> best practice, or is d<DEL id="blah1">i</DEL><INS for="blah1">o</INS>g permissible, and is it any of our business? "

<inserted> ScribeNick+ Steven

<Steven> Roland: I find it hard to get excited about this

<Steven> Roland: Further down it shows a reply, and mentions the edit="moved"

<Steven> Roland: Are there any more of these issues?

<Steven> Shane: Lots

<Steven> ... such as 82 issues in "incoming"

<scribe> Scribe: Gregory_Rosmaita

SP: meant to be new working group when renamed XHTML2
... HTML5 WG didn't take over our HTML issues
... we (XHTML2) retained issues from old working group
... done our best to work with them, go to LC, and let them comment on that

RM: should delete or mark incomplete in issues list
... can close one-by-one, but need to do SOMETHING with them
... new ones since group re-chartered?

SP: would be in incoming bucket

SM: only 1 issue since rechartered
... about editorial error in CURIE spec which i think i fixed

RM: comments on XML Events get to issue tracker?

old issue tracking system on XML Events: http://htmlwg.mn.aptest.com/cgi-bin/xhtml2-issues/XMLEvents?user=guest;statetype=-1;upostype=-1;changetype=-1;restype=-1

shane's tracker "XMLEvents has 6 messages"

<Roland> http://lists.w3.org/Archives/Public/www-html-editor/2008JanMar/0015.html

Errata for XML Events: An Events Syntax for XML

comment: In "B. Schema Implementation" the third sentence reads: "It is divided into an attributes module and an element module for the XML Events module defined in this Proposed Recommendation." This should read: "It is divided into an attributes module and an element module for the XML Events module defined in this Recommendation."

SM: don't remember seeing it in issue tracker, but can put it there
... didn't imagine us doing an update to XML Events 1

RM: should we ever need to, this is something we should address at same time
... should keep track in relation to XML Events 1 - if do need to PER XML Events 1, then would address all errata on that document

SM: i'll get it into the tracker

Subtopic: How Do We Make the Issue Tracking System Useful?

SM: use issue tracker effectively for new work (last call comments, etc.)
... in the past, have tracked comments from last call or PR review - works to extent we use it
... do have a lot of old baggage and unprocessed stuff in old tracking system
... can take all of unprocessed stuff and mark obsolete

RM: proposed resolution: All items in issue tracking system dated prior to new charter be closed as no longer relevant

SM: will that fly?

SP: can argue it reasonably; old group's charter ended and new group chartered (XHTML2) - did best to process as many comments from previous group as could and are now staring tracking afresh
... TBL did say this is going to be a new group with a new charter and new chair

SM: going to create a bucket called "old" -- everything that isn't in a certain state will be included
... if closed, leave where is - already dealt with and useful for pointers
... if implemented, done it, but haven't reissued public draft since effecting changes
... if marked "suspended" we were holding for later processing
... anything not in "suspended", "implemented" or "closed" will be obsoleted

RM: charter start date?
... end date 31 December 2009

SP: date of charter 2007-02-02

SM: anything before 2007?

SP: yes

RM: yes

GJR: yes

AC: yes

SM: should we inform users that comment tabled, please re-review and re-submit if necessary

RM: let's ascertain how many there are to deal with

SP: [counting]

proposed RESOLVED: All items in issue tracker dated prior to issuance of new XHTML2 Charter (2007-02) will be closed as no longer relevant

SM: more general question: 259 messages that are "open", but 135 are in the "trash" - spam not yet marked as closed
... rather than doing in "real time" will go through and remove the extraneous "crust"

SP: if only 80, can do quick pass over them
... can we chop it down even further?

RM: shane, take a pass, post a list, and will revisit if we need resolution

<scribe> ACTION: Shane - post list of outstanding old issue tracking system items to public-xhtml2 in preparation for final disposition [recorded in http://www.w3.org/2009/04/23-xhtml-minutes.html#action04]

<trackbot> Created ACTION-79 - - post list of outstanding old issue tracking system items to public-xhtml2 in preparation for final disposition [on Shane McCarron - due 2009-04-30].

D - a single element to indicate dialogue

<ShaneM> brb

SP: elements for structural portions of document; semantics indicated by attributes (why added role)
... in case of D, looks like adding semantics
... can't we use a role="dialogue"

GJR: would role="dialogue" be applied to Q

<Tina> If we are not indicating semantics with elements, then why are we indicating semantics with /some/ elements?

GJR: Q inappropriate because when including dialogue, one is not quoting, but speaking in the character's voice

http://www.w3.org/MarkUp/xhtml2/wiki/ProposedElements/D

"3. D provides a clear and necessary semantic marker in a generic/foundational declarative markup language -- it provides a declarative means of identifying text as dialogue, rather than leaving indications of dialogue embedded in prose to be indicated with "naked" quotation marks, character entity values or unicode values such as " or u0022 or u0027, and so on) nor can one rely on extremely spotty implementation of the :before and :after pseudo-elements to g

RM: dialogue in my understanding involves multiple parties

GJR: monolog would be subsumed into dialogue

http://www.w3.org/MarkUp/xhtml2/wiki/ProposedElements/D

<alessio> ...what about d role="monolog"?

RM: could we make this a property of P or SPAN?

MG: yep

what about this use case:

<d>

<dl>

<dt> Shopkeeper</dt>

<dd> <nd>[obviously lying]</nd> Sorry, we're just closing for lunch</dd>

</dl>

</d>

<Tina> Can't that be solved by pulling in another namespace specifically created for, say, plays etc?

"since there are many different contexts in which D is an appropriate element/container, a dialogue element demands a role or type attribute, in order to enable differentiation between types of dialogue; a very preliminary list of predefined role values for D follows: conversation, dialogue (covers theater, film, fiction, where the element is used to distinguish dialogue from other types of content, transcript "

SM: not convinced D/dialogue is a semantic structure

RM: INS/DEL kept only in text; use other attributes (including role) to annotate
... content model on wiki page is horrendous

"Note: If it is decided that two distinct elements are needed -- one for inline use and one for block use -- then the simplest solution would be to use D inline (since it would be used quite often in a document instance) and DB or DBLOCK for block use. Since dialogue is dialogue, however, it would be preferable to have a single element to declaratively mark dialogue, and leave the rest to native structural elements and stylesheets. Therefore, throughout this

RM: what is rationale?

"Rationale: XHTML2 currently lacks an element that can clearly mark text as dialogue. The use of Q is not an option for this use case, as the characters aren't being quoted, they are engaging in a conversation. Since a dialogue element would be used quite often, the shortest element name is the best element name for the semantic concept of "dialogue". "

http://esw.w3.org/topic/PF/XTech/HTML5/Dialogue

RM: represents fact that individual is uttering statement

GJR: provides a means to represent text that is expressed by an individual or group of individuals

RM: perhaps should make distinction --

<d>

<dl>

<dt> Shopkeeper</dt>

<dd> <nd>[obviously lying]</nd> Sorry, we're just closing for lunch</dd>

</dl>

</d>

RM: don't think D would be right thing there - DL for scene
... what if in argument or debate?
... generic D would be much more about annotating a section
... D doesn't carry semantics for me there

"when used as a block level element, there needs to be a child element of D which can declaratively mark part of the text as not being dialog, such as stage directions, cues, description of non-verbal action, and so on; in the following example, ND ("not dialogue"), has been used and is being proposed to fill this need"

<d role="transcript>

<dl>

<di>

<dt> Donegan:</dt>

<dd> Did you have any conversation with Brothman to the effect that

this information was going to the Russians?</dd>

</di>

<di>

<dt> Bentley:</dt>

<dd> I remember once he asked me if the Amtorg people l1ked what

they were getting.</dd>

</di>

<di>

<dt> Donegan:</dt>

<dd> Did you ever identify yourself to him as one who had worked for

the Amtorg?</dd>

</di>

<di>

<dt> Bentley:</dt>

<dd> No, I never had. I was introduced to him as his Communist Party

superior, and he was to give me dues and information.</dd>

</di>

<di>

<dt> Donegan:</dt>

<dd> Did he give you his Communist Party dues?</dd>

</di>

<di>

<dt> Bentley:</dt>

<dd> Yes, he did.</dd>

</di>

<di>

<dt> Donegan:</dt>

<dd> On how many occasions, would you say?</dd>

</di>

<di>

<dt> Bentley:</dt>

<dd> That would be hard to say. It wasn't every week, but it might be

roughly every month.</dd>

</di>

</dl>

</d>

DL Should Give Structure to, Not Generically Indicate, Dialogue

RM: investigate what people are actually saying -- all about speech

<h> Kane Found in Love Den with <em class="em-quote">Dancer</em></h>

em-quote would be a :before and :after psuedo-element to indicate quotation nesting

<Roland> <u role="to-be-said">Hi</u> <q role="actually-said">Hello</q>

<markus> Id be in favor of supporting inline D with the rationale to avoid misuse of Q, and wait for (along Tinas suggestion above) a future dedicated module under XHTMLMOD2 for "real" block level markup of plays etc.

<markus> ... Steven: you cant attach rdf/a to unicode quotes

MG: idea with D is can ascribe from whom dialogue originates

RM: Q says "this is what was actually said" - what does D do?

<Steven> The rule was for deciding how to introduce new elements

<Steven> we left old elements mostly untouched

<Steven> for 'mindshare' reasons

GJR: D proposal has role/type proposed for D

SP: understand use case, but feel that if our current markup extensibility methods are not able to handle this case, then somehow we failed
... next time someone comes along with a semantic element idea, what do we do? extensibility has to be considered before new elements are added
... how to fix this? semantic mechanism so that anyone can add semantic info to document
... if not sufficient, what is wrong with model?

RM: agree with Steven -- concentrate on problem statement
... if agree on problem statement, can explore what we can do within XHTML2's over-arching philosophy -- if run into problems, then can consider new element

GJR: semantically there IS a difference betwen a quote and dialogue

RM: role="maleProtoganist"
... take GJR's use cases and investigate if can solve them with existing technologies native to XHTML
... content model is scary

GJR: next steps?

RM: rationale is most important bit;

"5. since there are many different contexts in which D is an appropriate element/container, a dialogue element demands a role or type attribute, in order to enable differentiation between types of dialogue; a very preliminary list of predefined role values for D follows: * conversation; * dialogue (covers theater, film, fiction, where the element is used to distinguish dialogue from other types of content); * transcript "

RM: concentrate on problems then propose solutions
... FIVE MINUTE WARNING
... any urgent business

SP: need to ensure roadmap reflects reality

Next Steps on XHTML2

RM: goal is to get new draft out in May 2009
... can we create draft for publication in may?

MG: how does this correlate with PLH's induced discussions with SamRuby?

SP: have to carry on according to charter whilst addressed in w3c channels - wouldn't be addressed until end of year

RM: editors comfortable with may 2009 for publishable draft?
... what help needed?

SM: ready to publish now

SP: WG task -- assign and read sections

MG: what about XForms?

SP: just reference XForms - import XForms 1.1 with the following elements
... ok if in working draft as is
... XForms WG discussed attribute names yesterday - going to deprecate old attribute names and introduce new - old content will still work, but XHTML2 can go ahead and just use new attributes

RM: implications?

SP: clashes with XForms attributes -- target, for example

RM: is there a definitive list?

SP: thought solved for XML Events 2, need to check XForms
... will post pointer to XForms minutes
... very accomodating

RM: schedule another vFtF soon
... will discuss logistics of next virtual face2face at next week's regular XHTML2 WG telecon

<ShaneM> there are now 45 issues that are !Implemented

<scribe> meeting: XHTML2 Virtual Face2Face

meeting+ XHTML2 Virtual Face2Face

meeting- XHTML2 Working Group Teleconference

meeting+ XHTML2 Virtual Face2Face

Summary of Action Items

[NEW] ACTION: Gregory - create wiki page for hreflang and content-negotiation best practices and examples [recorded in http://www.w3.org/2009/04/23-xhtml-minutes.html#action03]
[NEW] ACTION: Markus - send email to public-xhtml2 about DAISY's use of CURIEs to public-xhtml2 [recorded in http://www.w3.org/2009/04/23-xhtml-minutes.html#action02]
[NEW] ACTION: Shane - get RDFa task force to contribute to, sign onto or issue CURIE comment on Last Call draft of OWL2 [recorded in http://www.w3.org/2009/04/23-xhtml-minutes.html#action01]
[NEW] ACTION: Shane - post list of outstanding old issue tracking system items to public-xhtml2 in preparation for final disposition [recorded in http://www.w3.org/2009/04/23-xhtml-minutes.html#action04]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.135 (CVS log)
$Date: 2009/04/23 16:08:24 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.135  of Date: 2009/03/02 03:52:20  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/Date:/Date - /
Succeeded: s/SP/SM/
Succeeded: s/can we use LINK/can we use "ref" with LINK to accomplish this?/
Succeeded: s/href/ href/
Succeeded: s/not/note/
Succeeded: s/what would be the Role value/what would be the rel value/
Succeeded: s/CURIE/CURIEs/
Succeeded: s/local/generic/
Succeeded: s/SP: can mention RIF/RM: can mention RIF/
Succeeded: s/RDFa implementation/rdfa.info/
Succeeded: s/SM: want to integrate/SP: want to integrate/
Succeeded: s/SM: can we use native/SP: can we use native/
Succeeded: s/SM: ITS copied ours/SP: ITS copied ours/
Succeeded: s/SM: can we use LINK/SP: can we use LINK/
Succeeded: s/rel="itsRule"/rel="itsrules"/
FAILED: s/rel="itsRule"/rel="itsrules"/G
Succeeded: s/SM: and ruby/RM: and ruby?/
Succeeded: s/hypertext attribute selection/hypertext attribute module/
Succeeded: i/Roland: I find/ScribeNick+ Steven
Succeeded: s/XML/HTML/
Succeeded: s/attac/attach/
Found Scribe: Gregory_Rosmaita
Found ScribeNick: oedipus
Found Scribe: Gregory_Rosmaita
Default Present: Steven, Roland, Markus, Gregory_Rosmaita, ShaneM, Alessio, Yves_S
Present: Steven Roland Markus Gregory_Rosmaita ShaneM Alessio Yves_S
Regrets: Rich_S
Agenda: http://www.w3.org/MarkUp/xhtml2/wiki/2009-04-FtF-Agenda#2009-04-23
Got date from IRC log name: 23 Apr 2009
Guessing minutes URL: http://www.w3.org/2009/04/23-xhtml-minutes.html
People with action items: - email force get gregory markus rdfa send shane task

WARNING: Input appears to use implicit continuation lines.
You may need the "-implicitContinuations" option.


[End of scribe.perl diagnostic output]