XHTML2 WG Weekly Teleconference -- 10 Sep 2008

<Steven> Hey Roland!

<Steven> wouldn't mind, or would mind, oedipus?

wouldn't

<scribe> Scribe: Gregory_Rosmaita

<scribe> ScribeNick: oedipus

<Steven> You're amazing Gregory. When I have a migraine, I mind everything!

<Steven> Exvcept lying under a duvet in a darkened room

survival tactic -- residue of nerve dammage

<Steven> ha

<Steven> Remind me where you were Roland

Previous: http://www.w3.org/2008/09/03-xhtml-minutes.html

Agenda Shaping and Announcements

SP: search for MarkB - sent 2 posts to emailing list in last 24 hours
... shall i chair

RM: please do

SP: TPAC registration: 6 weeks to go
... still room for presentations on TP day for those who wish to propose a presentation
... RDFa is now a PR - congratulations and thanks to shane
... Tina, from M12n acknowledgements as "Greytower Technologies"

TH: correct

SP: GJR as "invited expert"

GJR: correct

SP: timeline - 1) Schema Review - would like at least shane to be around
... markB sent review to list - http://lists.w3.org/Archives/Public/public-xhtml2/2008Sep/0006.html

http://lists.w3.org/Archives/Public/public-xhtml2/2008Sep/0007.html

Schema Review

SP: MarkB seems to only have positive review

RM: no feedback, just acknowledgement that read and reviewed

TH: want to re-read next iteration

SP: deadline for comments is this friday

<Steven> http://lists.w3.org/Archives/Public/public-xhtml2/2008Aug/0044.html

SP: asked to discuss specific items mark thinks we need not worry about
... decided not to get too worked up about assertions
... think we can just send confirmation and thanks for asking us to review

RM: sounds good

TH: plus 1

GJR: plus 1

RESOLUTION: send Schema confirmation that XHTML2 WG reviewed, thanks for opportunity, no comments

M12n Status

SP: shane -- only thing left acknowledgements?

SM: not ready to publish yet

SP: when?

SM: thought yesterday, so probably today

SP: ping me when ready and will send off necessary email
... ready to get stamp of approval after making sure draft is in final shape

XML Base Comments

SP: comments should have gone to both XForms and XHTML - reply only to XForms

http://lists.w3.org/Archives/Public/public-xhtml2/2008Sep/0005

<Steven> http://lists.w3.org/Archives/Public/public-forms/2008Sep/0015

SP: comment 1: clarification of URI - XML Base relationship to CURIEs

<Steven> http://lists.w3.org/Archives/Public/public-forms/2008Sep/0018.html

SP: reply suggests that it is up to XML Vocabulary to decide what is URL - good answer, but need to say so in spec; emailed asking if could just state declaratively -- seem to have said "yes"
... added comment which answers our comment perfectly well; accept?

SM: what does this mean for RDFa?

SP: if we say it applies, then it applies

GJR: plus 1 to that

SP: a CURIE should end up being relative URI once pre-processing done

SM: in grammars that accept XML Base

SP: yes, of course
... good answer - just say thank you for doing this
... just received reply - why not accept w3c position on what constitutes a URI - conflict between syntax space and value space
... don't mention value space, but that is answer we want -- answer is just "yes" if CURIE allowed as URI, then Base applies

RM: plus 1

<Tina> +1

GJR: plus 1

SP: comment 2: accepted (add example)

<Steven> http://lists.w3.org/Archives/Public/public-forms/2008Sep/0013.html

<Steven> http://lists.w3.org/Archives/Public/public-forms/2008Sep/0014.html

SP: comment 3: missing definition - comment about change list - URI reference change in RFC number, but not referenced in text
... replied that it is only in the references
... not sure value of reference only in references, but on other hand is harmless

RM: can waste some time figuring it out - put it in, so what is reason it is there?

SP: look at spec again

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

http://www.w3.org/TR/2008/PER-xmlbase-20080320/

SM: clear reference

SP: ok

<Steven> http://lists.w3.org/Archives/Public/public-forms/2008Sep/0015.html

SP: commment 4: existing applications may break
... may have been misunderstanding on our part - thought changing def of URI in Base (what is allowed in XML Base attribute) - was this change allowed to make to make PER rather than cycling through LC
... response: never case XML Base values are URIs - change is not a normative change - change in reference

RM: seems ok response to me

SP: no other issues to reply to

<scribe> ACTION: Steven - reply to XML Base comment replies [recorded in http://www.w3.org/2008/09/10-xhtml-minutes.html#action01]

Tina's Primer

<Tina> Zakim: unmute Tina

TH: background: freenodes web channel - started writing XHTML primer that is value neutral vis a vis values and principles; been writing up, took time but last draft done yesterday
... meant to be introduction to XHTML - explain where fits into web of today without taking political stance either way - suggests way of doing content negotiation, a bit of history and bit of detail; will end up in topic of #web channel at freenodes
... happy with it - more comments appreciated

SP: all should review it

<Tina> http://www.dev-archive.net/articles/xhtml.html

<scribe> ACTION: Working Group - review Tina's XHTML primer [recorded in http://www.w3.org/2008/09/10-xhtml-minutes.html#action02]

TH: already passed by shane
... need to revise with a bit about schema -- how to use to define if element can be child
... all comments good or bad, direct to me

<Tina> tina@greytower.net

TH: hold back on publication for a week so can make changes

RM: where published?

TH: dev-archive -- took over css.nu (CSS info site) -- publishing documents there that aren't related to any specific company - neutral place to publish articles and documents
... will stay at URL but not yet in index or atom feeds (http://www.dev-archive.net/articles/xhtml.html)
... genesis: flame wars over XHTML - a lot of misunderstanding;

SP: good work
... part of spearhead action to undo some of the dammage done to XHTML through misinformation and misunderstanding
... W3C at TPAC last year, presenter said 0.0% pages on web using XHTML -- spun the data - applications not served as appllication/xml but text/html when comes down pipe; analysis of pages on web found approximately half announcing themselves as XHTML; would be good if can make some announcement of that - 15% of top 20 web servers serve XHTML to undo some of the dammage
... even had to disabuse TBL of XHTML as failure canard

RM: what is it and why? a lot of pages not valid - why claim XHTML - what looking for in XHTML?

SP: part of problem is UAs don't validate, so no message it is wrong; like a compiler - same attitude to web pages - chuck at browser and if works as intended, everyone is ok

RM: BBC site comes out of often malformed server side

TH: if send as XHTML this will happen, if send as text/html this will happen; a lot of use of XHTML as HTML which results in poor pages

SP: hard to get feedback that it is wrong

TH: need to explain that need to know what is doing with XHTML; XHTML Transitional doctype is being treated as HTML4; all authors know is use XHTML

SP: similar to unicode - if character set is utf-8, many think have done their work, which isn't the case

TH: point of article - need to know what you are doing when using XHTML - here is what you need to know
... any comments, please send to me -- be as critical as necessary -- going to keep neutral

SP: if not valid, it is not XHTML - no, is incorrect XHTML, but still XHTML;

TH: problem with sending XHTML as HTML, XML parser doesn't get near it -- in that context it is HTML

SP: as long as intention of page is adhered to, don't serve things as XHTML for browsers, but for XML pipeline that allows XML output at end

RM: what are people's primary motivation? source serving PoV or delivery PoV

TH: dev-access uses XHTML
... transform XHTML using XSLT on dev-access
... most people don't need XHTML to start with - large educational problem involved
... have to get people to stop saying XHTML is evil - when used for purpose for which it was designed

[fyi] Open Accessibility uses XHTML as normative format (so can support ARIA and RDFa)

TH: a lot of people want absolutes

SM: XHTML mime discussion?

SP: yes

XHTML Mime Type

SM: have comment from simon peiters...

TH: link to comment?

<Steven> http://lists.w3.org/Archives/Public/public-xhtml2/2008Sep/0000.html

SM: very long comment

SP: summarize how we should deal with comment

SM: what is obligation - have to respond, but not address or satisfy all comments if cannot be satisfied?

SP: have to check process on notes

TH: can we simply thank him?

RM: ought to take them on board

TH: take points aboard

SP: wrong about RFC 2119 terms -

SM: talking about document, not abstract

SP: says "this abstract sucks. it shouldn't use RFC 2119 terms"

SM: document not normative, so nothing should be normative is basis of comments, i believe

SP: dusting off to reflect experience with UAs knowing what to do with XML; summary should say "should" because is quote from specificiations
... first comment i disagree with

SM: next comment: "not normative" don't reference RFC 2119 -- remove and use non-RFC 2119 terms

RM: if have no reference to RFC 2119, than "should" is just plain english

<Tina> The question is: does the use of RFC 2119 references *do any harm*? Does it in any way CHANGE the content?

SP: note's strength is that abstract contains capsule of note
... since SPieters took trouble to comment, should reply in good faith and positively

SM: from process persepective, for me to go through point-by-point, suggest resolutions and bring back to WG for discussion
... M12n Rec is priority

<scribe> ACTION: Shane - review SimonP's comments point-by-point, suggest resolutions and bring back to WG for discussion [recorded in http://www.w3.org/2008/09/10-xhtml-minutes.html#action03]

SP: need to wait for reply to comments until move forward

SM: receive any others?

SP: no, but did point HTC to it

RM: HTC call on friday

Role Module & Comments

http://www.w3.org/MarkUp/2008/ED-xhtml-role-20080625/

SP: have we dealt with all the comments on Role?

SM: requested transition meeting on Role a while ago

SP: will chase that down

SM: resolved to request CR on 9 july 2008 - cited in today's agenda

SP: follow up on that

SM: dependency on CURIEs was one lingering issue

CURIEs

SM: did anyone follow discussion on CURIEs in RDFa task force -- jonathan rhys sent in comments on 30 August 2008 to RDFa task force (not copied to public-xhtml2)

<Steven> http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2008Aug/0137.html

SP: quick ping issue?

SM: yes

<Steven> Thread starts here: http://lists.w3.org/Archives/Public/public-rdf-in-xhtml-tf/2008Aug/0133.html

SM: issue is JRhys believes need to define in CURIE spec the transformation fucntion to get from curie lexical space to curie value space - requirement of XML Schema Datatypes, so must be done; concrete comment against CURIEs need to resolve before request CR transition

RM: only for base types, not derived types

SP: quotes from post

SM: required telecon to get this resolved in RDFa TF

RM: built-in types?

SM: comments about CURIE draft not RDFa

SP: lexical space of CURIE is well defined syntaxically
... transformed to URI by either sticking the prefix and postfix bits together (concatinating to form URI) - result must be in any URI

SM: don't say that

SP: we don't?

SM: one place say has to be URI in another an IRI
... also say value space is IRI
... isn't XML Schema Datatypes 1.1 IRI is same as URI

SP: thought IRI was syntaxical/lexical space and URI is value space
... IRI cannot go over wire

RM: described in URI/IRI spec

SP: CURIEs transformed to IRI - when IRI gets sent over wire has to be transformed into URI

SM: out of our hands

SP: lexical comes from second transformation

RM: transformation occurs in circumstances other than over-the-wire -- other cases where should be transformed from IRI to URI

SP: IRI defines relationship
... what does RDF expect? URIs or IRIs?

SM: expects URIs
... doesn't anticipate existence of IRIs
... they are tokens, so almost doesn't matter; IRIs are lexical space in real world, and there is not a 1 to 1 mapping from IRI to URI - not isomorphic - many to one mapping -- more than one IRI representation
... URIs are subset of IRIs
... subtle angles-on-head-of-pin stuff -- wont' get this from discussion
... if way to make clearer to get from lexical to value space and requirement of XML Schema than should take comment on

SP: think i understand comment -- assumed good enough to say concatonate together and form an IRI, but surprised CURIE spec doesn't say that

SM: it does

SP: so what is problem?

SM: not expressed in terms of XML Datatypes - no machine way to express concatonation
... may be why in TAG some have argued for new datatype schema for this
... is a tuple - doesn't concatonate

SP: minute over -- need to go to another call

RM: promised TAG by end of september

SM: won't rush but please cogitate upon this

ADJOURNED

- DRAFT -

XHTML2 WG Weekly Teleconference

10 Sep 2008

Attendees

Contents