Meeting minutes
light attendance today, will keep light minutes to match
TPAC hybrid event: W3C needs a feeling of how many people hope to be in person or know they can't be there. Survey due by March 28.
we report our availability to attend in person and Janina and Mathew will report back for APA.
And the usual: which groups do we need to talk to and why.
Agenda Review, Membership & Announcements
Action Items
https://
https://
if we can get some of our use cases (like education/testing) to appeal to smart speakers and implementor technology that's not covered by the ARIA approach, we need to highlight that.
I think it's possible to extend the key/value pairs beyond locale and into things like voice
janina: HTML lang may not be sufficient for variations in dialect/vernacular/historical pronunciation
<janina> http://
I think the only way to address this in SSML is with specification of a voice (assuming the voice pack reaches all areas of the dialect requested).
<janina> Koine Greek is need for New Testament as well as Septuagent (Jewish) studies
Alan: is there a fallback in the ARIA approach
<janina> That's different But Koine is inappropriate for Plato and Euripides
Mark: it helps cases of sub but not say-as or pausing which are primary concern of this group.
Mark: math and chemistry requires pausing between terms
Mark: currently, it's hacked with commas to create the pausing which affects the Braille output.
Mark: say-as numeric could be keyed into the ARIA approach with IPA
aria-pronounce="as-characters" could wrap several tokens and each has a key for the locale
I think we could do this for our 5 primary features
janina: we may want to approach them before we do all the mapping if it's just a mapping exercise
Mark will complete the task at hand as well as try to draw attention to the 5 primary features we're requesting during an initial implementation
I will write up a treatment of the expanded ARIA approach and Janina will carry it forward for further discussion with ARIA WG at a later date.