See also: IRC log
<scribe> last minutes are at http://www.w3.org/2006/07/18-i18ncore-minutes.html
Francois: last meetings minutes - any comments?
... nothing
<scribe> ACTION: Everybody to look at input mode discussion (ONGOING) [recorded in http://www.w3.org/2006/07/18-i18ncore-minutes.html#action01]
Felix: I had a short look, what should we do now
Richard: It will come up in the XHTML review again
... but other people might have input here
Francois: let's keep it ongoing
<scribe> ACTION: all to look at charmod resid test suite and gather feedback (PENDING) [recorded in http://www.w3.org/2006/07/18-i18ncore-minutes.html#action02]
<scribe> ACTION: Francois to build a current issues list on charmod norm (PENDING) [recorded in http://www.w3.org/2006/07/18-i18ncore-minutes.html#action03]
<scribe> ACTION: Mary will make the updates discussed in Mandelieu on ws-i18n (ONGOING) [recorded in http://www.w3.org/2006/07/18-i18ncore-minutes.html#action04]
Mary: working on that
... should I just start a second document on the policy?
Francois: Felix asked what remains in the first document
Mary: it would describe how the headers look like
Francois: why having two documents?
Mary: since you might implement the stuff with a different framework
Francois: it is more difficult to do more than one document
Felix: it might help the adoption to have a separate document, which would be in line with the other policy
domains
... Mary, could you edit both documents?
Mary: it depends, we have to see
Francois: could we start with one document, and see what happens
Mary: o.k., let's see if we need two documents
Francois: the action item is ongoing?
Mary: Yes
<scribe> ACTION: Richard to create a wiki on "first letter" (DONE) [recorded in http://www.w3.org/2006/07/18-i18ncore-minutes.html#action05]
Richard: not a wiki, though, see:
<r12a> http://www.w3.org/International/notes/firstletter.html
Richard: the current wikis are a pain to edit as they get long
... I have a hybrid between a wiki and a
mailing system
... Karunesh, do you think that is useful?
Karunesh: The current way will work
... I will make some comments
Richard: sent them to www-international
Francois: very nice, but more work for Richard
<scribe> ACTION: Richard to review XHTML Basic 1.1 and XHTML Modularization (ONGOING) [recorded in http://www.w3.org/2006/07/18-i18ncore-minutes.html#action06]
Francois: Mark is not here
Felix: I'll ping him again
Francois: we cannot go on without Mark
<scribe> ACTION: Felix to ping Mark on LTLI discussion [recorded in http://www.w3.org/2006/07/18-i18ncore-minutes.html#action07]
Felix: how about matching - should LTLI discuss matching, not normatively?
Richard: if it is not normative, it should not be in LTLI
Felix: it was just brought up a while ago
Francois: LTLI could identify use cases for matching on the web
... that is more BP than normative spec
Richard: not sure
... this could slow things down for LTLI
... this could slow things down
Francois: that's a worthy point
Felix: summary: people at the call currently think explanatory material on matching should not be on LTLI
Francois: that's all for this week
http://www.w3.org/TR/2005/WD-ws-i18n-20050914/
Francois: already handled during action items
<fyergeau> http://lists.w3.org/Archives/Public/public-i18n-core/2006JulSep/0004.html
Francois: discussion state: we need use cases for LTLI
Francois describes a use case which uses the browsers language
Richard: we agreed that xml:lang will not be used for local, right?
Mark: no
... you could either deny xml:lang
... or is see it as the core of locale, which also gets
other information from somewhere else
... the latter seems to be more practical
... "locale" has no well established meaning like
language
... e.g. for some people, locale definitions requires at least three countries
... e.g. the nationality of a user, the place of
transaction, the place of the user etc.
... "locale" is an amorphous set of preferences
... I think, the core part is language
... so
people would add other fields for their requirements
Richard: should we say "a core for locale"
... or "the key indicator"?
Francois: both
... it is core, but you can use it also as an indication
Richard: that is: the other aspects of the locale
Mark: take a look at time zone
Francois: what is the use case for time zone on the web?
... to give time indications in the users time
zone?
Mark: yes
... e.g. for news to say "this happened three hours ago" or "at two o'clock"
... though
sometimes you don't do that
Francois: time zone is the from language most independent locale-part, so its outside the core
Mark: the same holds for currency
Francois: for most languages you can make a good guess
Richard: what is "a locale"?
Francois: we already agreed that it depends on the application
Richard: so the locale is not something you can label?
Mark: the notion of locale is very amorphous
Francois: you want locale identifiers so that a recipient of some information is able to do s.t. with it
...
it is a collective thing
... it applies to a group of people
Richard: one could say "a locale groups together types of information and values that defines the interaction in
a particular situation"
... but we refer to the usage of a particular label
Francois: we need to distingush locale identifiers versus their meaning in different contexts
Felix: we said two weeks ago that locale is always related to processes (interactions), whereas language is a kind of static
Richard: I wanted to say "xml:lang provides a component of the locale information"
Mark: in most cases the core of the locale is provided by 3066
... if you look at Java
... posix adds a
charset to the notion of locale
... but the core is really the language
... there is variance in the industry
... but not for the core
of locale
Richard: "the core of the locale is given by RFC 3066bis"
... that is: locale is a larger thing, and
language is only a part of it
... as you think about locale identification
Mark: I don't think that we should be too precise in the language
... that is: saying that xml:lang provides
the core of the locale
... for many applications that is sufficient
... you just want to cover the 99% cases
Richard: I liked the way Mark phrased it
Francois: posix locale is a problem
... you may need various ISO standards. We have to make sure that this
work
[discussion on posix versus other locale identifiers between Mark and Francois]
Mark: if you specify the country
... you need additional information
Richard: could someone summarize where we are now?
Francois: I could try, but I would be biased
Richard: we could set up a wiki
<scribe> ACTION: Richard to create a wiki for the LTLI discussion, Francois to make a summary of the current discussion [recorded in http://www.w3.org/2006/07/18-i18ncore-minutes.html#action08]
Felix: charmod, I'll bring it up on the mailing list