Hello Marcos,
many people from the i18n core WG are away this week, so there might be
more replies later. This is a personal reply.
Marcos Caceres wrote:
Hi, i18n-WG.
In recent feedback we received from Addison Phillips regarding the
Widgets 1.0: Packaging specification, he suggested that WebApps should
add a <span>-like element to our Widget Configuration Document format
(so to allow bidi text to be declared).
I think such an element would only be necessary within these elements:
name, description, author, license. It seems that only these elements
may contain human readable text.
At our last F2F, WebApps discussed the proposition and we were left
wondering if we can use unicode's RLM/LRM characters instead of a
<span>-like element? Can i18n please advise us on this?
See
http://www.w3.org/TR/2008/NOTE-xml-i18n-bp-20080213/#DevDir
and
http://www.w3.org/TR/2007/NOTE-unicode-xml-20070516/#Bidi
I will not repeat the arguments here, but the conclusion is that indeed
an attribute for directionality information would be better than relying
on Unicode control characters.
Not having the
<span>-like element significantly simplifies our processing model. We
don't want to sacrifice i18n for the sake of simplicity, so we really
need your guidance again on how to move forward.
I personally would recommend you to use the <its:span> element in the
ITS namespace. The element is defined at
http://www.w3.org/TR/2007/REC-its-20070403/#span
This element gives you the "dir" attribute and various other attributes
which are useful for esp. Widgets localization. See
http://www.w3.org/TR/2007/REC-its-20070403/#att.local.no-ns.attributes
See also the related "Best Practice" to define such an element for XML
vocabularies at
http://www.w3.org/TR/2008/NOTE-xml-i18n-bp-20080213/#DevSpan
To keep simplicity for Widgets 1.0, you could say in your conformance
description that a Widgets processor has various options to deal with
the <its:span> element (or more in general: the ITS namespace) and its
attributes: ignore them or process them.
If you do not want to add markup from a specific namespace, you could or
should IMO add extensibility points for people who need such markup.
That is, change in the schema something like
description = element description {
xmllang.att?,
text
}
to
description = element description {
xmllang.att?,
any
}
and define "any" and a pattern "anyElement" as
any= (attribute * { text }
| text
| anyElement)*
anyElement = element * { any }
Again the conformance for such markup can say: ignore it ("it" meaning:
markup from other namespaces) or process it. I think you are basically
saying that already at http://www.w3.org/TR/widgets/#extensions
Regards, Felix.
Having read "Internationalization Best Practices: Handling
Right-to-left Scripts in XHTML and HTML Content", we are aware that
there are problems with text editors ATM, but we are hoping the tools
will improve as Unicode support becomes more common place (or is that
wishful thinking?).
Kind regards,
Marcos
