<scribe> Scribe: Addison Phillips

<scribe> ScribeNick: aphillip

<David> Hello all

Minutes and Agenda

Action Items

all: review HTML5 (rolling action item to be reviewed weekly)

addison: invite representative from SSML to discuss pronunciation alphabet registry for next time

<scribe> DONE

update draft to support option C for specifying preferences in ws-i18n

write a couple of simple examples of issue #3

<scribe> DONE

respond on our behalf to CSS on ruby issue

Pronunciation registration (SSML)

(addison summarizes)

burn: spoke to ltru "back in the day"
... they said "not clear what is different from LTRU tags, but might be somethign different"
... effort made to not duplicate that work
... some concrete examples of things that wouldn't appear in ltru's registry
... one is IPA
... another would be alphabets such as arcabet or worldbet, which are alphabest used in speech synthesis for many years
... simplifications of IPA
... additionally, we expect that there are organization (companies, univerisites,etc.) that may experiment by creating their own
... and we want to enable that
... these are the most concrete examples of what would appear in pronunciation registry but don't belong in bcp47
... examples are for pinyin and jyutping
... and only pinyin defined
... but that's because that is most needed
... two tags: pinyin and pinyin2001

http://lists.w3.org/Archives/Public/public-i18n-core/2010JanMar/0004.html

<burn> http://tools.ietf.org/html/draft-burnett-pronunciation-alphabet-registry-00

richard: clarify two points
... people might want to experiment with codes; would they registry the experiemental codes?

burn: depends on whether they wish to standardize the codes or not
... x-code is private use
... or non-x-hyphen could be registered
... need for expert review to prvent stuff like registering something called "IPA" that was something else

richard: didn't say why not down language subtag route, so assumed that it would be easier to deprecate items in this registry?

burn: language subtag registry contains mostly broad consensus subtags
... whereas our registry might have a very small audience

zh-p-pinyin2001

und-p-ipa

burn: couldn't use the pl alphabet separately
... we provide only alphabet for production... purely for production
... codes could be used by only a single TTS vendor
... most TTS systems are either diphone or triphone systems
... if I wanted to make my complete set of underlying models available, I might also want to register it
... so that authors could access it
... not convinced that's the best use, but people might have different opinions
... goal is strictly pronunciation
... language subtag registry more focused on understanding

fsasaki: various hepburn subtags registered
... might be overlap?
... also, impose a burden on two registries to coordinate their work?
... might coordinate registrations to avoid having different strings/tags for the same purpose

<r12a> http://rishida.net/utils/subtags/index.php?find=hepburn&submit=Find

burn: (clarify) if you want to say "if in ltru then not permitted in pronunciation", then have a real problem

fsasaki: did not want to imply that, but just avoid having different subtags in each registry for the same purpose

<r12a> http://rishida.net/utils/subtags/index.php?find=pinyin&submit=Find

burn: also, need a specific list of what is supported/not supported
... not a broad list of subtags of which some might apply
... reviewer could object to "that's not enough detail for me to know what that means"

addison: maybe have it both ways--both extension and standalone registry?

tlh-Cyrl-AQ-hepburn

http://tools.ietf.org/html/bcp47#section-3.7

burn: would you be comfortable with this if we did the extension? need a general feeling

addison: I agree; I think you're doing something different and it might be useful to also have a 'p' extension

richard: both politically and practically useful, I think

<fsasaki> +1

burn: may ask of you to join our group when presenting

addison: I'd be happy to come help you

HTML ISSUE-88 content-language-multiple

http://lists.w3.org/Archives/Public/public-i18n-core/2009OctDec/0106.html http://lists.w3.org/Archives/Public/public-i18n-core/2010JanMar/0002.html http://www.w3.org/International/wiki/Htmlissue88

richard: we've talked about this before
... html5 currently says that content-language pragma (i.e. META)
... is limited to one language
... which is inconsistent with previous
... we talked at TPAC, richard wrote a chagne proposal based on our conversation with Hixie
... proposal basically says comma-separated list
... if no lang attribute, then the <meta> tag could be used to infer document language, as long as it only has one language
... if more than one language then you can't infer the document processing language
... and then you can look at http header
... similarly
... so what we have there is a precendence path

lang attribute -> meta -> http header

<scribe> chair: any opposed to this text?

none opposed

addison: looks good to me
... concern about harmonizing terminology

richard: document wide default language
... propose to replace with language declaration pragma
... or maybe pragma language

addison: I like this and like the text you've proposed

<scribe> chair: (resolved) go forth and promote this on our behalf

AOB?

<r12a> i have made two changes to http://www.w3.org/International/wiki/Htmlissue88#Impact

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/extensino /extension /
Succeeded: s/pragme/pragma/
Found Scribe: Addison Phillips
Found ScribeNick: aphillip
Default Present: Dan_Burnett, aphillip, David, felix, Richard, [Oracle]
Present: Dan_Burnett aphillip David felix Richard [Oracle]
Regrets: Mati
Agenda: http://lists.w3.org/Archives/Member/member-i18n-core/2010Jan/0002.html
Got date from IRC log name: 13 Jan 2010
Guessing minutes URL: http://www.w3.org/2010/01/13-core-minutes.html
People with action items: