<trackbot> Date: 28 January 2009

<scribe> scribe: Steven

XHTML Media types

Steven: It should be up any minute this afternoon

http://www.w3.org/TR/xhtml-media-types/

@lang

Roland: If we want that in the specs, then we would need the change before we do the transition requests

Shane?: I don't mind

scribe: but it only really affects 1.1

Roland: Mobile Web want it in Basic as well

Shane: @lang is harmless and helpful

Roland: And there is support from I18N

Shane: We have to say something about how the two interact
... but putting @lang into 1.1 invalidates the XHTML Media types doc
... Someone from Opera complained about us minting new DOCTYPEs
... because they hardwire the DTD names
... I told them that we have rules for DTD names, and it ought to work
... so if you serve Firefox a 1.1+RDFa doc, then character entities don't work

Steven: But the media type says it

Shane: True
... But I am now nervous about telling people to use new DOCTYPES

Tina: People need character entities
... Tricky

Steven: This is a bug in the browser
... strictly they should recognise the DTD
... if they take a shortcut
... recognising a small set of hardwired DTD names is nto good enough
... they should use the namespace (or something)

Shane: Or the media type

Roland: Aren't we off topic; this is a different issue to @lang

Shane: Yes, I ratholed
... the fundamental issue isn't to authors who deliver as text/html, but application/xhtml+xml
... we need to address the larger issue for 1.2

Roland: Anybody opposed to adding @lang to 1.1 for this transition?
... and Basic

RESOLUTION: add @lang to 1.1 and Basic before transitioning

Steven: Should I continue with the transition of 1.0, or wait for them all to go together?

Shane: I can make the change really quickly
... I will send out tomorrow proposed language
... until Friday for comment
... Transition request Monday

Steven: OK

Tina: No objections

XHTML2

Tina: I prefer to keep useful elements in

<Roland> INS and DEL http://lists.w3.org/Archives/Public/public-xhtml2/2009Jan/0012.html

Steven: The attributes are much more useful
... you can't delete or insert a table row for instance

Roland: I think that doing this as metadata works quite well
... but if people are used to ins and del, we could keep them as well

Shane: It is difficult to express the content models of ins and del with schema and DTD

Steven: Attributes make more possible, and in the bargain solves the problem of the messy content models

Roland: insertion and deletion is metadata about a paragraph for instance, not structure

Tina: But what about included text

Roland: That was my proposed compromise
... only allow ins and del in text content

<alessio> +1

Roland: Let's discuss this in the list then

p content model

Steven: I don't see what the problem is
... we enlarged the content model to include what people think of as a paragraph
... People sent us use cases that we agreed to support
... you don't *have* to include quotations or tables in a paragraph, but you have the option

Tina: I'm not convinced that they are valid use cases
... can we dig out some language expert who will resolve this?

Roland: We shouldn't legislate what the user may think of as a paragraph
... the author may decide

Shane: At the moment you can have a p inside a p
... I checked the issues case and indeed we have people asking for this
... but also the reverse

<ShaneM> I am now convinced by Steven's example.

<ShaneM> Steven's exeample was a paragraph that said "Today I will do: (some table of stuff) then I will go home."

Steven: I want to be able to write "Here is an example: [[1][2][3][4]], which show sthat...."

<alessio> I've seen many example, expecially in the public administrations sites

Steven: and I don't want to be disallowed that.
... If you don't want to do that, don't do it, but please don't try to stop me doing it.

<ShaneM> p content model is currently (PCDATA | Text | List | blockcode | blockquote | pre | table )*

Roland: Time's up
... I will study the content model further

[Adjourn]

<alessio> bye! :)

This is scribe.perl Revision: 1.133  of Date: 2008/01/18 18:48:51  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/Sahne/Shane?/
Succeeded: s/min/mind/
Succeeded: s/bosy/body/
Succeeded: s/adding/to adding/
Succeeded: s/mail/list/
Found Scribe: Steven
Inferring ScribeNick: Steven
Default Present: Roland, ShaneM, mgylling, +04670855aaaa, Tina, Steven, Alessio
Present: Roland ShaneM mgylling +04670855aaaa Tina Steven Alessio
Regrets: Gregory
Agenda: http://lists.w3.org/Archives/Public/public-xhtml2/2009Jan/0072.html
Found Date: 28 Jan 2009
Guessing minutes URL: http://www.w3.org/2009/01/28-xhtml-minutes.html
People with action items: