See also: IRC log
-> http://www.w3.org/XML/XProc/2015/09/02-agenda
Accepted.
-> http://www.w3.org/XML/XProc/2015/07/08-minutes
Accepted.
No regrets heard.
-> https://ndw.github.io/specification/langspec/small-fixes/head/xproc20/
Henry: I have a pedantic
suggestion, in 2.10.2, in the first new paragraph, it says
"this section describes how..."
... There is no corresponding paragraph in 2.10.1.
... I think there should be.
Norm: Sure, I think that's a good idea.
Henry: I'm still distracted by the contradictory statements that AVTs are attributes and strings, but fixing that is probably not worth it.
Alex: Do we say that it's an XPath expression?
Norm: I can check that.
Henry: Last time we talked about
this, we talked a lot about the interaction between media type
and charset.
... I think there's an implicit flowchart here. Would it help
to summarize the order of what you're going to find out before
it starts?
... The order of play is: if there's an encoding, the string is
decoded. After that you have a byte sequence and don't care
about the encoding anymore.
... If there's a charset param, you then construct a string
using either the charset parameter or UTF-8
Norm: Yes
Henry: I think this "flowchart" should be the the second paragraph of the section.
Alex: Generate a sequence of bytes, etc.
Henry: If the media type is a
non-XML media type, and there is no charset parameter, what is
used to decode the byte sequence.
... I think we can say UTF-8, unless it's XML, in which case
the XML decl wins.
Alex: If it's a text media type,
utf-8.
... The situation about the XML decl never arises, because
we've made that an error.
Norm: We should say utf-8 is the default.
Henry: It might be useful in the flow chart to just have almost footnote numbers cross-referencing the errors.
Alex: In 5.10.1, non-XML is
implementation-dependent.
... Are we open to being able to treat any kind of
semi-structured data as something you can run expressions
across?
Norm: In principle, yes.
Alex: Can you do non-XML in implicit inlines
Norm: Not today.
<jfuller> changes look good to me
Henry: I still don't understand
what it means to remove in-scope namespaces.
... And there's a typo in 5.10.2.
Some discussion of removing in-scope namespaces.
Alex: We have a different situation; sometimes we hand documents to a tool.
Norm: We could have used exclude-result-prefixes but it would have to be used everywhere and that might lead to conflicts.
Proposed: accept this doucment as the status quo, with the understanding that the flowchart will be added.
Accepted.
<jfuller> note I redrafted errors proposal
<jfuller> https://github.com/xproc/specification/issues/136
<jfuller> for next week
Thanks, Jim.
Some discussion of possible future f2f meetings.