Internationalization Core Working Group Teleconference

23 Feb 2011


See also: IRC log


Addison Phillips
Addison Phillips


trackbot, prepare meeting

<trackbot> Date: 23 February 2011

<scribe> Scribe: Addison Phillips

<scribe> Scribe: Addison Phillips

<scribe> ScribeNick: aphillip

<r12a> hurrry up zakim

<r12a> kenny, are you joining the call

Minutes and Agenda review

review documents for review and such

Action Items

Change the language tags article to say that now there is an extension defined, and describe it briefly

trackbot, close ACTION-4

<trackbot> ACTION-4 Change the language tags article to say that now there is an extension defined, and describe it briefly closed

also did "choosing language tags"

please send comments

<r12a> http://www.w3.org/International/questions/qa-choosing-language-tags

Check on HTML5 language tag validator and send any comments

addison: need to copy list

<r12a> http://www.org/International/track

Update review radar and ping www-international seeking comments

close ACTION-8

<trackbot> ACTION-8 Update review radar and ping www-international seeking comments closed

Review techniques index pages to check that all the current links in the index point to the right place

Publish language tags article

trackbot, close ACTION-22

<trackbot> ACTION-22 Publish language tags article closed

Propose advancement of UTR#20 to publication based on our revision and noting that we plan a substantive revision

addison: done... will check why not in archives

Info Share

richard: Unicode version 6 complete text was published

<r12a> http://www.w3.org/blog/International/2011/02/22/unicode_version_6_0_complete_text_of_cor

richard: multilingual web workshop in Pisa
... anyone who would like to speak, please send in registration
... or anyone who wants to come

<r12a> http://www.multilingualweb.eu/documents/pisa-workshop/pisa-cfp

CSS3 transform "capitalize" thread


addison: propose that we ask Unicode to assemble this data?

<kojiishi> http://dev.w3.org/csswg/css3-text/#text-transform

kojiishi: just updated the spec, defining it as each word in title case per Unicode
... word definition is unclear
... and internationalized titlecase is difficult
... saying in following text that UAX29 suggested and implementation defined

<kojiishi> 'tis -> 'Tis

kojiishi: word segementation could be better defined for capitalization
... also provide for language dependent rules and tables

richard: capitalize option depends on finding each word and then capitalizing
... but also word or phrase done sometimes?

fantasai: just did an update
... copied over text
... put first letter (grapheme?) in titlecase
... even if you go back to CSS1, doesn't do anything language specific
... everything straightforward except for definition of "what is a word"
... points to UAX29 for text segementation
... this is not the way to do proper title case
... rarely gives good results
... just define well enough to make it work reliably

r12a: you've changed word to ttitlecase

fantasai: should be checked in

r12a: is there an option to capitalize beginning of phrase
... not just whole phrase

fantasai: have to do sentence detection
... stuff like embedded spans would interefere

r12a: sort of like "first letter" though

fantasai: need to correct text to "titlecase" and fix tests instead of uppercase

r12a: first-letter is more useful than each word
... may still be language specific issues
... may be useful to ask Unicode to spec out more

fantasai: does user-agent recognize (e.g.) ' or T as first letter in 'Tis
... does UAX29 recognize? probably just use for text segmentation
... then apply to each word




kojiishi: looking at spec
... maybe say use Unicode definition
... and not "first letter"

<kojiishi> http://www.unicode.org/versions/Unicode6.0.0/ch05.pdf

<kojiishi> 5.18 Case Mappings, Titlecasing

kojiishi: section 5.18,
... apply Unicode titlecasing to each word

addison: allow for language specific?

fantasai: want to see it consistently broken in all cases

r12a: if we point to Unicode, then up to Unicode how far they take it

fantasai: is this worth having inconsistent implementations?

r12a: if we point to Unicode, then now it will be consistent (ly bad)
... but could improve over time
... might be able to plug in specific languages

fantasai: can't count on it generally, even if work on your computer
... can't count on results is not helpful

norbert: don't agree with word "style", since this changes text

r12a: articles on our site written in sentence case, but using a style to lowercase them
... so it really is styling
... proper e.g. German casing should be in text
... this is just "prettyfication"

fantasai: feature doesn't work most of the time and won't
... really need human intervention
... or sophisticated linguistic analysis
... and already making undefined

Norbert: in German, casing makes a difference in meaning
... e.g. fliegen vs. Fliegen
... not just matter of style
... titlecasing would change meaning

r12a: so you'd remove style for translation into German

Norbert: really a dangerous feature

addison: is it deprecated enough?

fantasai: did add some text, will take to CSS WG to ask them (if you want)

r12a: if I use this, would probably use in :first-letter case only
... so I would find it useful in that case

fantasai: can give example of words broken

<David> can you hear me?

<David> 1 minute while I redail

addison: couple of proposals:
... first, ask Unicode to assemble information
... second, ask CSS to further deprecate or provide negative examples

David: recommend deprecation
... a lot stronger in saying in some cases that it will alter meaning

r12a: thinking about this in cases where use style and then text translated
... but you wouldn't use this property in the translation
... something that works in English
... not so dangerous

<kojiishi> if you deprecate, you may also loose first-letter case as well?

r12a: put health warning, but most of time it will be perfectly obvious that it not working right

fantasai: can be useful with first-letter

<scribe> chair: proposals...

UNKNOWN_SPEAKER: ask Unicode/CLDR to work on additional language-specific tailrings and data for titlecasing

<r12a> yes

fantasai: aginst
... request when we actually need it

norbert: agree

kojiishi: could ask to optimize definition of word for titlecasing?
... improve UAX29


David: don't have a use case, significant work
... so not worth asking at this point
... ask for things we need

kennyluck: defer

r12a: got impression mark as keen to do this

fantasai: on their todo list, but not high priority

<scribe> chair: sense is that we're say "no formal request"

UNKNOWN_SPEAKER: UAX29 improvements?

fantasai: no strong opinion

<r12a> yes

r12a: these things need to be done one day

<scribe> chair: ask for formal deprecation? or ask for further negative examples, etc.?

r12a: the latter

norbert: can see first-letter case
... corner cases

<kennyluck> suppport

fantasai: will just edit into spec, just need clear understanding of what to put there

addison: have WG members provide comments and suggested text?

fantasai: yes, post to www-style@

what time is the call at?

David: looking like 1500 UTC

<David> http://doodle.com/b58wqx87g4hqgkbc


fantasai: proposal to hyphenate...
... is all hyphentation just on first line only

r12a: some scripts where hyphentation happens on second line
... ?

fantasai: need specific example

Summary of Action Items

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.135 (CVS log)
$Date: 2011/02/23 16:04:33 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
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)

Succeeded: s/first letter/:first-letter/
Found Scribe: Addison Phillips
Found Scribe: Addison Phillips
Found ScribeNick: aphillip

WARNING: No "Present: ... " found!
Possibly Present: Addison_Phillips David Norbert P12 P19 P28 Richard ScribeNick addison aphillip fantasai freedom kennyluck kojiishi matial r12a trackbot
You can indicate people for the Present list like this:
        <dbooth> Present: dbooth jonathan mary
        <dbooth> Present+ amy

Agenda: http://lists.w3.org/Archives/Member/member-i18n-core/2011Feb/0013.html
Found Date: 23 Feb 2011
Guessing minutes URL: http://www.w3.org/2011/02/23-i18n-minutes.html
People with action items: 

[End of scribe.perl diagnostic output]