MHonArc v2.5.0b2 -->
dita message
[Date Prev]
| [Thread Prev]
| [Thread Next]
| [Date Next]
--
[Date Index]
| [Thread Index]
| [List Home]
Subject: RE: [dita] Final xml:lang proposal
From: Michael Priestley <>
To: "Grosso, Paul" <>
Date: Tue, 4 Apr 2006 13:44:49 -0400
re:
>There is no contradiction. We've already
said that
>xml:lang inherits in the usual sense and more local
>settings take precedence over settings higher up
>(and we didn't even have to say that, since that is
>the only way xml:lang can work per the XML spec).
Existing documented behavior for metadata
in maps is that it adds to or overrides metadata in the referenced topics.
Maps are effectively closest to the deliverable, and thus are more likely
to be accurate about things like audience, platform, publisher, etc. But
xml:lang is acting in the exact opposite way, which makes sense, but I
would think is worthy of comment. I don't know of any other standards
that involve the pushing/pulling of metadata in the way maps do and (separately)
conref do, so I'd be surprised if the xml:lang spec says something explicit
about it.
Even if the xml:lang spec does accurately
and completely cover this behavior, I still think it's worth calling out
explicitly in our spec, since it is (deliberately) not consistent with
the behavior of other attributes. We would at least want to update the
spec to say something like "metadata in maps overrides metadata in
topics when there is a conflict, except for xml:lang, where the metadata
on the topic wins".
Michael Priestley
IBM DITA Architect and Classification Schema PDT Lead
http://dita.xml.org/blog/25
"Grosso, Paul"
<>
04/04/2006 01:31 PM
To
<>
cc
Subject
RE: [dita] Final xml:lang
proposal
While I believe we're all trying to say pretty much
the same thing, I'd like to delete the paragraph:
In the case of a contradiction between the xml:lang
value set on the map and the xml:lang value set on
the topic, the setting on the topic overrides.
There is no contradiction. We've already said that
xml:lang inherits in the usual sense and more local
settings take precedence over settings higher up
(and we didn't even have to say that, since that is
the only way xml:lang can work per the XML spec).
This gratuitous paragraph makes it sound like there
could be a "contradiction" and that we might be
suggesting some behavior for xml:lang other than
the XML-spec-defined behavior. There can't be
(a contradiction) and we aren't (specifying some
other behavior), so we should just delete these words.
Also, I'm not sure about the wording:
When reading XML markup that embeds scripts of different
languages, the embedded languages ....
I don't think scripts can be in an XML document. An
XML document consists of characters (and markup). Besides,
the first half of the sentence talks about embedded scripts,
and the second have says embedded languages. Also, it doesn't
seem to make sense to "When reading...when the document is
saved."
I think we mean to say:
When an XML document contains strings of characters from
different languages, those strings should be indicated
via markup.
but I could be wrong. In particular, I'm not sure if we're
trying to say that editor tools should convert any use of
the Unicode direction-switching characters into markup.
paul
>
