See also: IRC log
No regrets heard.
Norm: We've got a bunch of open issues. Henry, want to give us an update?
Henry: The only changes I've got planned are the one's minuted last week. Renaming of the profiles, renaming B' to S, and changing of a bit of wording in 4.2.3.
Norm: Thank you, Henry.
... Let's skim over the open issues and see how many we feel comfortable marking as closed.
... I think issue 1 is resolved.
Vojtech: I think so.
<scribe> ACTION: Henry to review isssue 3 (2) and 3 (3) as editorial making changes as the editor sees fit [recorded in http://www.w3.org/2011/10/13-xproc-minutes.html#action01]
Henry: WRT comment 4, add a
sentence to the beginning of section 3 clarifying how the
classes are derived from the table below.
... Comment (5) is resolved by section 7 in the latest draft.
Norm: That leaves (1); Liam objects to calling the Infoset a data model.
Henry: If I changed it to "the infoset or data models that capture similar information"...is that OK?
Norm: Issue 4 was a thread we did about browsers. I think we can close that.
Alex: It's possible the browser
could do more, but I don't think we're going get a better
... I'd like XInclude, but we don't have a profile for that the browsers could do.
Henry: I think it's fine; in so far as the point of the full profile is that you basically have the document that the author committed to, I think having XInclude but not external entities is odd in that respect.
Alex: Unless you're in a DTD-less world, then it's not odd.
Norm: Issue 5 is about test cases, but we're going to try not to do CR, so we can close it, yes?
Norm: Issue 6 from the Core
... I think we've removed "recommended" which helps, and I think we'll need something in the introduction to spell out why we chose the profiles we did.
<scribe> ACTION: Norm to attempt to draft that prose, cf. cmsmcq's comment 1 [recorded in http://www.w3.org/2011/10/13-xproc-minutes.html#action02]
Norm: Issue 7; I don't think we're going to get the browsers to move on that.
Alex: The point there was the second bit, having standalone have an interpretation.
Henry: The browsers aren't going
to pay attention to the standalone declaration.
... Unless we change the XML spec to change the default. The problem is that the default is standalone=no. So if we ask the browsers to change to make standalone=no an error, we'll break all XHTML. It's a lose-lose situation.
... The one thing we could imagine doing is to say that there's a media-type dependent default which is standalone=yes. What we'd be asking the browsers to do is two things: (1) give an error in the presence of an explicit standalone=no, and (2) give an error for non-HTML XML unless there's an explicit standalone=yes
Norm: In 1997, maybe. But today it's just not worth it. We'd be asking every user serving non-XHTML XML to change.
Henry: So how would Core feel
about saying that the XML XHTML5 spec can default
... If we don't do this, then we should have raised an issue on XHTML5 saying that they're not raising an error when XML says they should.
Further discussion, leading to the observation that standalone is a validity constraint
Paul: I'm happy to have the Core
WG say something if it helps make things work better.
... as long as it doesn't rewrite the XML spec.
Alex: I think the question is, if you look at the combination of our new document with the smallest profile and the XHTML5 spec, what's the interpretation of the standalone attribute.
<scribe> ACTION: Paul to put standalone on the Core agenda. [recorded in http://www.w3.org/2011/10/13-xproc-minutes.html#action03]
Norm: Let's skip 8 for the
moment, I think we've resolved 9 by removing the word
... I think 10 is resolved.
... I think 11 and 12 are just observations, not comments on the spec
... I think 13 is resolved by adding section 7
<scribe> ACTION: Henry to add a note to the effect that we are talking about static parsing, not dynamic environments [recorded in http://www.w3.org/2011/10/13-xproc-minutes.html#action04]
Henry: issues 14 and 15 are informational, not comments on the spec
Norm: Issue 16 is clearly a bug.
Norm encourages everyone to read xproc-dev