See also: IRC log
<scribe> Scribe: Addison Phillips
<scribe> ScribeNick: aphillip
<scribe> Agenda:http://lists.w3.org/Archives/Member/member-i18n-core/2010Jul/0000.html
aharon: DOM3 keyboard events
<Norbert> http://trac.tools.ietf.org/wg/httpbis/trac/ticket/214
norbert: http content-language
addison: forward date-time proposal to HTML WG
<scribe> ACTION: addison: blurb the -u- extension last call [recorded in http://www.w3.org/2010/07/14-core-minutes.html#action01]
richard: publish the b and i tags/elements document
<scribe> DONE
richard: send our comments and questions to the html wg on polyglot document
<scribe> DONE
<r12a> http://unicode.org
richard: unicode is now
cool
... they tweet and fb and rss and such
... and they did their first such announcement yesteray
richard: forwarding to our list
http://lists.w3.org/Archives/Public/public-i18n-core/2010JulSep/0018.html
aharon: basically the proposal is
to add another field (attribute) to DOM3 key events
... and also text events
... which are new to DOM3
<scribe> ... new attribute would indicate the language of the keyboard
UNKNOWN_SPEAKER: would be a string (locale), in BCP 47 format
aharon: use case: smart
quotes
... differ by language so user-agent could convert based on
input keyboard
... and on-line editors might want to do the same thing
... basing on the UI language is not necessarily the document
language
... bidi use case is...
... word looks at keyboard language to decide if neutral
character such as a number or punctuation
... to decide if part of ltr or rtl string
norbert: active spell checker
might be a better choice
... no need to switch layout/input locale to type various
languages
aharon: keyboard language could be useful hint
richard: ability to turn off, etc.
addison: input locale may be more
than just the language, includes other info, such as input mode
or layout??
... warnings about usage might be important
aharon: generate markup or
unicode controls could be generated based on this to control
bidi? provides an option for this
... doesn't necessarily impact html req doc
mati: somewhat
uncomfortable
... use cases are different things
... smart quotes are language dependent, for example (and
language => locale in many cases)
... so smart quotes make some sense perhaps
... but bidi use case is not locale dependent
addison: input locale != system locale
mati: not sure indicator for
first use case is same as what's needed for second case
... mixing things together
aharon: usually don't indicate what language I'm typing, but still need to type documents in various languages (he, ru, en, etc)
mati: but might use same keyboard
for en, fr, de
... don't have to switch keyboards to type for example a lot of
latin script languages
... it's a cultural preference connected to current content
language, not the keyboard
aharon: system locale is not indicative there either
mati: in hebrew layout you can
input latin (ltr) text from it too
... I don't think it's a good indicator for smart quotes
norbert: matter of
documentation... here's a facility, insufficient in and of
itself, but can be useful
... interesting to see if there is another use case in which
the input language can be used more specifically
addison: maybe change speller if the key event indicates a keyboard change?
<scribe> ACTION: Aharon: draft text proposal for keyboard language/input locale attribute for DOM3 key events [recorded in http://www.w3.org/2010/07/14-core-minutes.html#action02]
http://trac.tools.ietf.org/wg/httpbis/trac/ticket/214
norbert: was in a meeting with
mnot a few weeks ago
... discussing content language
... he wrote a ticket, so what do we want in HTTP spec
(quoting) Some people have interpreted the "of the intended audience" as meaning that this header indicates the language of the user, not the language of the content. As a result (for better or worse), current practice in the i18n community is to ignore this header (although there are other reasons, e.g., the lack of accuracy in the event of multiple languages being present). Proposal: remove "the intended audience for"
richard: aligning with HTML @lang
doesn't make sense
... intended as metadata, not as document processing
language
... think it's a bad idea
norbert: mnot@ thinks it might be editing error, but you're saying this is intentional?
richard: yes, it is a way of
specifying information about langauges in document as a
whole
... the intended audience
norbert: is anyone using as intended?
richard: when content negotiation happens, it is filled in based on accept-language (etc)
norbert: any user agents use this?
richard: there are some browsers
that start to interpret
... such as choosing fonts for CJK, or :lang styling
addison: TMS or CMS systems
can/do use it??
... not necessarily broken?
richard: some implementations do
odd things, such as assign multiple langauges to all content in
a document
... let's keep these things distinct and clear and
separate
... roy fielding
... look for Issue-88 in html archive for examples
<scribe> chair: do we want to take a position on this?
norbert: proposal: don't remove "intended audience" and make clear that it is not document processing language
addison: think that http today says something like that (not the same as indicating all languages used in content)
<scribe> chair: any objections to this being our position?
<r12a> [[
<r12a> The docs at
<r12a> http://www.w3.org/International/questions/qa-http-and-lang
<r12a> are correct. It would be best if HTML5 didn't break them
<r12a> and all of the systems that depend on those semantics being
<r12a> part of the Web architecture.
<r12a> ]]
<Aharon> I am in favor
norbert: +1
addison: +1
richard: +1
none opposed
<fsasaki> fine by me too
<najib> Me too
<scribe> ACTION: Norbert: update ticket for httpbis to indicate our WG's position [recorded in http://www.w3.org/2010/07/14-core-minutes.html#action03]
<r12a> [[
<r12a> The HTTP headers refer to the entire representation. If the
<r12a> representation is intended to have an audience of multiple
<r12a> languages, as is often the case when side-by-side translation
<r12a> is desired or mandated, then the content should be labeled
<r12a> appropriately. That use case is often found in government
<r12a> documents, poetry, lieder, language lessons, dictionaries, etc.
<r12a> I would expect HTML content to be tagged as a single language,
<r12a> if any, at some element level, whereas meta and link should
<r12a> support multiple languages at the resource or representation
<r12a> level.
<r12a> ]]
This is scribe.perl Revision: 1.135 of Date: 2009/03/02 03:52:20 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Found Scribe: Addison Phillips Found ScribeNick: aphillip Default Present: +972.7.474.6aaaa, +44.141.888.aabb, Richard, David, aphillip, +972.5.225.5aacc, +1.415.283.aadd, matial, Norbert, Aharon, [IPcaller], Felix Present: +972.7.474.6aaaa +44.141.888.aabb Richard David aphillip +972.5.225.5aacc +1.415.283.aadd matial Norbert Aharon [IPcaller] Felix Agenda: http://lists.w3.org/Archives/Member/member-i18n-core/2010Jul/0000.html Got date from IRC log name: 14 Jul 2010 Guessing minutes URL: http://www.w3.org/2010/07/14-core-minutes.html People with action items: addison aharon for httpbis norbert ticket update[End of scribe.perl diagnostic output]