Norm apologizes for messing up GMT in the agenda. And for the long delay since our last meeting

Accept this agenda?

-> http://www.w3.org/2010/html-xml/2011/06/28-agenda

Accepted.

Accept minutes from the previous meeting?

-> http://www.w3.org/2010/html-xml/2011/03/08-minutes

Accepted.

Next meeting: 12 July 2011 (in two weeks)

Accepted.

Review of comments

-> http://www.w3.org/2010/html-xml/2011/06/comments/

Norm: I closed comment 001 as it wasn't really a comment.

Comment 002 from David Carlisle

Norm: I think these are all editorial and am happy to accept them. Any disagreement?

None heard:

Comment 003 from Julian Reschke

Norm: I'm of two minds on this one. I think of XML as having a DTD language and a set of vocabularies.

Henri: I think Julian's point is well taken, I suggest we call it a framework.

Norm: I'm happy to take that as editorial direction and make a stab at it.
... Any objections?

None heard.

Comment 004 from Noah Mendelsohn

Norm: Isn't really acomment...

Comment 005 from Robin Berjon

-> http://www.w3.org/2010/html-xml/2011/06/comments/#C005

Norm: The first two are editorial.
... Robin suggests we just say "polyglot has limited applicability"

Henri: I think Robin's comments all make sense.
... I'd say that the the parser doesn't actually guarantee that the output is valid, only that it's well defined.

John: Tagsoup is the same way.

Norm: Ok, if there's general agreement about Robin's comments, I'll attempt to address them all.
... Any objections?

None heard.

Comment 006 from Henri Sivonen

Henri: The first comment is about the fact that the draft is misleading. It doesn't "just work". It's way overoptimistic. There are issues with, for example, empty tags.

Norm: Perhaps I overstated the case. I'll try to soften that statement. Or at least make it more accurate.

Henri: I suggest just removing the paragraph.

Norm; That looks reasonable to me.

Henri: The substance of my next comment matches the substance of one of Robin's.

Norm: Ok, I'll try to fix that as well.

Henri: My next comment is about interfacing subsystems. Using the HTML parser when you have XHTML5 in a larger document is just weird. I would assume that you wouldn't clip anything out and hand it to an HTML5 parser. You'd pass the parsed object model around, not the text.
... I don't think the HTML5 parser needs to be involved in any way.
... And then there's something about MIME and such and I think that steps outside the scope of using an XML document as a container.
... It seems out-of-place.

Norm: That was explicitly called out as one of the possible solutions when we were considering the use case.

Henri: What the document says about the script tag is factually incorrect.
... what you get in the DOM is not escaped in any way.

Norm: Yes, thats's a totally misleading statement. I'll fix that.
... Any other questions or comments about Henri's comments?

None heard.

Comment 007 from Kurt Cagle

Norm: I think this is about the same paragraph Robin and Henri commented on.
... Does anyone have any suggestions for the illustration Kurt suggests?

None heard.

Henri: I think there's a slippery slope towards rewriting the polyglot guide if we go that direction.
... I think we can provide some examples without trying to be exhaustive.

Bob expresses a desire for some examples.

Yeah, sorry Yves, I borked the agenda because of DST

Comment 008 from Noah Mendelsohn

Norm: I think Noah and John have both observed that making XML "more forgiving of errors" is a harder problem than the draft suggests.
... I'll try to improve that.

Any other business?

Adjourned.