Internationalization Core Working Group Teleconference

Minutes and Agenda review

aharon: bidi update on html5
... lists and images
... validity of unicode controls in html

Action Items

Dump charmod-norm and add disclaimer paragraph, dates, etc. so that we can publish an interim working draft

Change 512 to 1024 in the charset declaration article

close ACTION-59

<trackbot> ACTION-59 Figure out how to keep alive open bugs for the html5 LC closed

close ACTION-62

<trackbot> ACTION-62 Write to HTML5 chairs and ask for extension to complete our comments targetting mid-August (??) closed

Info Share

HTML5 bidi

aharon: dir=auto...
... the implementation results in webkit
... I saw that it wasn't doing everything
... oversight in the specification
... concentrating on text nodes
... and has no meaning for inputs
... and no meaning for text-area
... opened bug on that
... and good response from bidi-aware folks
... no editor response yet
... hope Ian gets to it
... concerned that it will be overlooked
... particular issue with lists
... the CSS3 lists module
... has changes for bidi, but still in flux
... still in draft
... the problem is that it has an implication for html5
... because CSS2 default is obviously bad for bidi but cannot change in CSS3
... so HTML5 needs to specify in its default stylesheet the correct CSS3 items for bidi as default...
... but CSS3 can't be referenced due to its draft status :-(
... fantasai advises "don't worry"
... images...
... wanted a new feature for image flipping
... was in WD of CSS3 images
... but cut because no implementations
... but it was only in the draft for 3 months
... that feature will be moved to CSS4

addison: protest to css?

<scribe> ACTION: Addison: send email to style@ advising that we want bidi image mirroring in CSS3 images [recorded in http://www.w3.org/2011/08/10-i18n-minutes.html#action01]

<trackbot> Created ACTION-64 - Send email to style@ advising that we want bidi image mirroring in CSS3 images [on Addison Phillips - due 2011-08-17].

<scribe> ACTION: aharon: send addison details on CSS3 image mirroring [recorded in http://www.w3.org/2011/08/10-i18n-minutes.html#action02]

<trackbot> Created ACTION-65 - Send addison details on CSS3 image mirroring [on Aharon Lanin - due 2011-08-17].

aharon: unicode formatting characters in HTML
... not an action from us
... some features can be handled by the unicode controls (some folks suggested)
... I expained why this is bad
... filed bug to disallow unbalanced controls
... recently Ian put that into spec
... so: within an attribute such as title, formatting characters must be balanced and within all text nodes of an element
... problem is that the language also makes 2 problematic statements
... 1) characters allows except when unbalanced
... and adds note that "as a matter of convenience" can use dir, bdi, etc.
... my problem is first that we should permit but discourage
... discouraged period
... markup is not "matter of convenience" but the preferred manner
... reopened bug

<aharon> http://www.w3.org/Bugs/Public/show_bug.cgi?id=11234

<scribe> ACTION: addison: add to bug 11234 that I18N WG favors the changes requested [recorded in http://www.w3.org/2011/08/10-i18n-minutes.html#action03]

<trackbot> Created ACTION-66 - Add to bug 11234 that I18N WG favors the changes requested [on Addison Phillips - due 2011-08-17].

HTML <rb> (i18n-issue-26)

<kojiishi> http://www.w3.org/Bugs/Public/show_bug.cgi?id=10830

Addison: should we reopen <rb> tag

richard: everyone in world wants this
... but ian is not swayed to provide this tag
... think this stance is not good for backwards compatibility
... and since IE9 and other browsers support
... asking for it to be optional, not for complete change to ruby model

<aharon> Can anyone summarize the reasons that ian is rejecting it?

<r12a> aharon, i think that's actually the problem - we can't see why

If this is just asking for <rb>, not for the feature that makes <rb> useful, then this is definitely a WONTFIX. Legacy IE browsers ignored <rb> in practice. In the model supported by the spec, the element provides no useful purpose.

richard: non-IE browsers didn't ignore <rb>

koji: number of <rb> cases cited by myself and fantasai

aharon: needs a bit more, since Ian feels ruby can be provided using other markup
... the thing to stress that <rb> serves useful purpose

richard: reopen and ask they address why our arguments are not good enough

<scribe> ACTION: addison: reopen 10830 for WG indicating that we are not satisfied, citing the usefulness, backward compatibility, and current implementation [recorded in http://www.w3.org/2011/08/10-i18n-minutes.html#action04]

<trackbot> Created ACTION-67 - Reopen 10830 for WG indicating that we are not satisfied, citing the usefulness, backward compatibility, and current implementation [on Addison Phillips - due 2011-08-17].

richard: propose the text to the WG first

addison: good idea

<scribe> ACTION: addison: propose text for reopening 10830 to WG before actually opening it [recorded in http://www.w3.org/2011/08/10-i18n-minutes.html#action05]

<trackbot> Created ACTION-68 - Propose text for reopening 10830 to WG before actually opening it [on Addison Phillips - due 2011-08-17].

koji: another bug about <rtc>

<kojiishi> http://www.w3.org/Bugs/Public/show_bug.cgi?id=13113

(all): need to read this

scribe: include next week

richard: fantasai suggested that jukugo ruby could be done via styling and I seem to remember thinking along the same lines at one point - also it may be possible to do ruby on both sides using nesting of ruby markup
... so maybe complex ruby isn't needed
... but I haven't yet been able to grind through the nitty-gritty details to confirm that.
... any progress on that front?

koji: no new information
... but ruby on both sides works better with complex
... although can be done
... but html wg is negative about complex ruby
... could live without

richard: solution is nesting ruby rather than complex
... tried and it isn't as bad as it could be
... complex is more than we needed, so would be happy if we could avoid
... but haven't checked if that is always possible

koji: hoping to make css-ruby spec progress
... hope to verify that all use cases are covered
... but haven't implemented yet

AOB?