I thought we had said we wanted to talk about xhtml media types document first today

(since we did not get to it last week)

<Roland> we did

okay. I think the mail message of interest is http://lists.w3.org/Archives/Public/public-xhtml2/2008Dec/0020.html

<scribe> Scribe: ShaneM

<scribe> TOPIC : Status of CURIE CR

Document should be published today or tomorrow.

Status of transition for PERs

There is a gating problem - we need to update the XHTML Media Type document first.

Shane thought we could request it anyway, but Steven explained that this will happen quickly so we should not do it until we are ready. Might be able to move on this depending on another agenda item. Tabled.

XHTML Media Types draft

We received a late comment: http://lists.w3.org/Archives/Public/public-xhtml2/2008Dec/0020.html

<Steven> Scribe: Steven

Shane: There a couple of comments here that are technical, but some are not
... these are issues for the W3C team to deal with
... we cannot respond to those parts
... he says that the doc talks about 4 media types, but really it is only about 2
... text/html and application/xhtml+xml
... the other two are not really needed

Roland: Sounds reasonable, this is about recommended usage

Shane: Good comment from this guy, as far as I can interpret
... He thinks we should rename the document to Media types for XHTML 1.x

Roland: Sounds right

Shane: Well Print and XHTML+RDFa wouldn't be covered then

Roland: It is now called XHTML Media types

Tina: Keep it as it is and say that it only applies to existing Recs

Steven: Notes are never normative

<alessio> agree

Steven: We should say that it just applies to existing recs

Mark: We should not just fight the battles at each step anew
... we should make this document applicable for generic cases
... the text/html mime type is about switching into HTML mode
... authors shouldn't be discouraged from that if it is a reasonable deployment strategy

Roland: There is more to this spec than just that
... it includes authoring guidelines

Shane: No, Appx A gives hints for any version

Mark: it is more political than technical
... the question is in my mind, is it OK to switch the browser to HTML mode, and since people are doing that anyway, it is

Tina: Shouldn't we just address the technical issues?

Steven: I suggest that we say that it applies to existing recommendations

<ShaneM> Proposed text: "The suggestions in this document are relevant to all XHTML Family Recommendations at the time of its publication."

<alessio> @shane +1

Roland: What do the authors who read this document want?

Mark: On the backplane XG there is a discussion about using Ubiquity to implement XHTML2
... well to do that you need to be able to use the text/html mime type

Shane: I proposed a sentence above

"The suggestions in this document are relevant to all XHTML Family Recommendations at the time of its publication."

Tina: I agree, keep the title, and add the sentence

<alessio> agree

Roland: I propose that too

So resolved

Shane: An early question asks us to say that it is only the opinion of the WG

Steven: No need for that. The information comes from recommendations

Shane: Anyway it says it comes from our WG
... He asks about text/xml and application/xml
... we don't care about those, right?
... he maintains that we don't need the space before trailing slashes on modern browsers.

Steven: True, but not to the point

Shane: He also complains that we are suggesting that authors produce invalid HTML4

Tina: He's wrong, it is valid

Mark: Anyway there is no agreed way to validate HTML4

Tina: Well, you can according to the DTD

Shane: Finally about named entities
... which may not work in generic XML parsers
... I don't think we care for this document
... Then comes the political stuff that has nothing to do with this document

Steven: The word XHTML does not appear in their charter

Shane: I request that Steven raise the issue of the use of the name XHTML within the team

Steven: Three is a domain discussion coming up

Shane: I propose we publish this document with today's changes

<Tina> +1

<alessio> +1

+1

<markbirbeck> +1

<mgylling> +1

RESOLUTION: Publish XHTML Media Types with topday's changes

<scribe> Scribe: Shane

<scribe> scribenick: shanem

<Roland> http://lists.w3.org/Archives/Public/www-html/2009Jan/0001.html

HTML 5 and XHTML 2 combined

There is a discussion thread started at http://lists.w3.org/Archives/Public/www-html/2009Jan/0001.html

markbirbeck: Benjamin has made some interesting comments that revealed existing browser behavior mark was not previously aware of.
... it seems that there are people out there who understand what we are trying to achieve with XHTML 2.
... there might be a way to make our message clearer. For example there might be confusion about what "deprecated" means. Can we help clear that up?

Roland: thinks this is an interesting thread.
... he points out that he cannot find anything in web space that describes the goals and philosophy of XHTML 2.

Steven: willing to take an action to draft a document.

<scribe> ACTION: Steven to write a draft of a short document on the philosophy of xhtml 2 for use in web space. [recorded in http://www.w3.org/2009/01/14-xhtml-minutes.html#action01]

<trackbot> Created ACTION-44 - Write a draft of a short document on the philosophy of xhtml 2 for use in web space. [on Steven Pemberton - due 2009-01-21].

<markbirbeck> http://internet-apps.blogspot.com/2005/01/xhtml-2-as-universal-document.html

<markbirbeck> :)

Roland: Is there a way to define a profile of HTML5 that maps to XHTML 2? And would that be valuable?

Steven: Does not think it is as difficult as it appears. We broke XHTML2 up into other bits so we could produce them separately. Completing XHTML 2 is just a matter of collecting those bits and we are done!
... He does not think we need to refer to HTML 5 in order to deploy XHTML 2. There is a client-side strategy via Ubiquity, and server-side strategies that involve transformation.
... "HTML is the assembly language that we are compiling down to."

Roland: The difference is that XHTML 2 is a very large language spec.

markbirbeck: HTML 5 has borrowed ideas from XHTML 2. That's fine. The bits that Mark finds frustrating is that they incorporate those ideas rather than referencing the modules that we have already produced.
... Basing XHTML 2 on HTML 5 seems circular.

Roland: we need to try to support our audience (paraphrase)

markbirbeck: We can't ignore how we got here. The modules are the basis of creating future languages and having extensibility. We can accelerate the process by using this architecture. A monolithic document makes that extension challenging.

Roland: profiling doesn't just mean subsetting. You can subset and extend.

<alessio> yes

markbirbeck: we could produce a document that describes the rendering behavior of the various html document types (?)
... The HTML 5 approach goes against the general modular philosophy that the W3C has pursued for years. People want to merge things together in ways that we will never imagine.

Roland: The thought process should be "what do we need to do to XHTML 2 to help ease deployment?" Are there things we should rethink based upon new information?

Steven: We have done this evaluation at times. There are things in XHTML 2 that are unavoidably different than HTML 4 / 5. XForms, XML Events

<alessio> completely agree

Steven: The HTML 5 approach to events is not scalable to the future, for example.

Roland: Can we take the time to evaluate where we feel that HTML 5 and XHTML 2 have common ground?

Will return to this next week.

ShaneM: and Welcome to the editing team, Markus!

<alessio> bye :)