W3C

– DRAFT –
(MEETING TITLE)

11 November 2024

Attendees

Present
alan, IrfanA, janina, matatk
Regrets
-
Chair
Irfan, Paul
Scribe
janina, matatk

Meeting minutes

Meeting Spoken Presentation Task Force Teleconference

Date 11 Nov 2024

<matatk> present_

Planning further work

matatk: Reviews the back story: We met with CSS at TPAC and had a great reception

matatk: They want to see CSS-Speech adopted.

matatk: Asked for tech agnostic use cases

matatk: We have come to believe CSS-Speech is a lot like "styling," and we can use a subset of it

matatk: Problem is implementations--how do we get some?

matatk: A need is well documented in community; Ed sector is using data- to accomplish

matatk: We have a notion on how to move forward ...

matatk: Our approach seemed to big ... Browsers were reticent to implement SSML

matatk: Plan: Document use cases as tech agnostic as we can

matatk: Goals from that would be distinguish homonyms

matatk: there are others

matatk: Suggested language lerning use case

matatk: APA Chairs and our Staff Contact Roy are in favor of this approach

alan: CSS Speech?

PaulG: More of a divide and conquer approach

PaulG: Some aspects in CSS-

-Speech; but an additional component

alan: Example?

alan: assume css speech for breaks, etc

alan: other would be content--but what? example would help

PaulG: Specific pronunciation

<janina> s/specific/Author-controlled specific

PaulG: We still need solution for abbrievations, subs, etc

PaulG: sub, alias ...

<Zakim> matatk, you wanted to talk about mappings

matatk: One advantages of this approach that we might be able to solve more fundamental problems, e.g.

matatk: styling might work on mappings; so perhaps emphasis is mapped into a11y-tree

matatk: or to ignore it!

matatk: We were considering sarcasm

matatk: If we can make the other module about semantics, it would be powerful

PaulG: How we say numbers is another big one

PaulG: The more accuracy the better

<Zakim> matatk, you wanted to ask about next steps, and groups

PaulG: Believe this two pronged is more likely to succeed

matatk: Feels like we have an emerging consensus for the two pronged approach

matatk: First step would be the use case doc

matatk: Next would be deciding what kind of group to work .. CG?

matatk: So much easier to get people into a CG

alan: Changing to css-speech would be a lot of work and a concern because of so little support from browsers

alan: We're trying to speak dialog boxes and suddenly some implementation of Web Speech API is pronouncing 'No' as 'number'

alan: There are good reasons to like css-speech, could have different speaking profiles

matatk: We're back to the beginning actually, but our experience makes us believe we can still get there

matatk: ?Real world experience documented would be helpful

matatk: Any experience on list using web api would be super helpful

PaulG: We will likely have multiple solutions, but we need some approaches that honor defined spoken presentation

PaulG: instead of heuristics!

PaulG: Do have hope after this last TPAC

PaulG: Language evolves fast, and heuristics (especially in the speech engines) don't. We need author control to ensure people are included.

matatk: data- is great -- pave the cow path

matatk: It helps us make the case that good standards need to be created

matatk: Any thoughts on how to start?

matatk: Emphasizes participation here is open and treasured

IrfanA: Should we start a new use case doc?

matatk: Let me take that under advisement ?

matatk: Needs to be use cases that will communicate. It may just be a reformulation of what's already there

matatk: We'll never lose what we have even if we make a new version

<PaulG> I'm in favor of talking about this in email or another meeting this week. I need to drop.

matatk: Suggest pick out use cases separate from tech

matatk: Suggests asking CSS with what we propose we could do. Phps split gap from use cases

matatk: Suggest we draft some before we ask for feedback not to overburden people

IrfanA: suggest creating a skeleton of new doc --a template

Minutes manually created (not a transcript), formatted by scribe.perl version 238 (Fri Oct 18 20:51:13 2024 UTC).

Diagnostics

Succeeded: s/allassume/alan: assume/

Failed: s/specific/Author-controlled specific

Succeeded: s/allother/alan: other/

Succeeded: s/suceed/succeed/

Succeeded: s/suddenly the OS/suddenly some implementation of Web Speech API/

Maybe present: PaulG

All speakers: alan, IrfanA, matatk, PaulG

Active on IRC: alan, IrfanA, janina, matatk, PaulG