Pronunciation Task Force Teleconference

15 Jan 2020


Irfan, JF, paul_grenier, janina, SteveNoble
Irfan Ali


<JF> Zal,im, who is here?

<Irfan> scribe: Christine

use case and gap analysis merge document

Review of document show gaps and repetitive infomation; we need to determine the approach as it appears we have derailed from the approach

In the approach we identify 5 examples, with information; in our explainer document we have focused on 2 approaches. In the gap analysis and use case, we are a little off track

There are 5 parts, with the first being the introduction. We have added mixed language scenarios and expectations, but this can be taken in phase 2 and not as part of the features we have identified.

In addition, rate, speech and volume be descoped for another version of the document i.e. phase 2.

We should focus the features we have previously identified and complete this document

Paul: to paraphrase, we should use language that are based on our use cases.

Irfan response - Examples we have provided in use cases are different from the core features.

Janina - should we put off language or not? We can always add declarative markup, which is not a discovery

Janina - we have a declarative mechanism and we are in an educational sphere and we may be limiting ourselves in many disciplines - medicine, music, law, etc.

Janina - keep declarative language and not detection of language

Janina - we should be specific in what we limit oursleves

Paul - the only thing that describes language is a lang attribute. In the context of language markup it is not detected.

Janina - the expectation is that we are pronouncing something that is marked up

Mark - I don't see the problem that Paul proposes

Irfan - is it good to provide a code example

Mark - more example always help

Paul - some examples may not require a lang attribute and sometime we don't want the pronunciation to make sense to the end user. What is then the recommendation for declarative? We need examples or guidelines for authors

<Zakim> Janina, you wanted to ask about lang

Irfan - changes to the document require discussion

WCAG requires that foreign language requires markup

The lang attribute declares the language and we augment on how it is pronounced

Are we looking to effectively mandate that you cannot use the pronunciation without the declarative lang attribute?

Janian - it does matter what the lang is for end users to know if they require language software

John - If using the pronunciation mechanism then you must declare the language, is that what we are asking?

Janina - we are only requiring how a word is pronounced

John - declare the expected language and then this is how it is pronounced, is that what we want?

Janina - what is the scoping extent of the pronunciation spec to look for alternative? Are we supporting this?

Mark - it would be great to have a perspective from TTS vendor on precedence.

Irfan - discussion with TextHelp, the implementation used is lang attribute for a particular word, or more specific a particular voice.

Mark - lang tag differs between specific speech synthesizers

Mark - we may not be able to solve this question.

Janina - we may need to add languages supported in the metadata.

Mark - example of current technologies detect language should be included in the document

Irfan - how to handle rate, speech and volume; can we add it in a separate section?

Irfan - summarizing the comments and come up with some agreement with some of the team members, and discuss in subsequent call/meeting

Irfan - Gap analysis section has empty sections, which we need examples to be added.

<Irfan> https://raw.githack.com/w3c/pronunciation/merge-doc/gap-analysis_and_use-case/index.html

use case issue triage

<Irfan> https://github.com/w3c/pronunciation/issues/50

Irfan - Paul has created a domain pronunciation

Paul - wanted to know whether we should pursue this, or set it aside. For someone that relies on speech to text to inform navigation, there is a security issue that uses phishing tactics

Paul - main concern is security and we can punt this to other groups

Janina - defer to w3c groups for guidance

Janina - maybe we exclude URIs

Irfan - add a comment waiting for review

explainer DRAFT document

Irfan - no other edits have been made to the Explainer document

Mark - we are in good shape, still in draft

Ready to be published?

Janina - this are published as a general explanation

Will follow up as there are no prescribed expectations; will follow up with APA

Roy - to publish Explainer there should be a w3c note and consensus

Irfan - format document to match other documents we have (i.e. w3c format)

Irfan - in agreement to publish, convert to required format and send

other business

Mark - we are targeted to be at ATIA conference and present the work of this group, as a call to action to the AT community to support this group

Irfan - APA working group meeting to include Leoni Watson

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes manually created (not a transcript), formatted by David Booth's scribe.perl version 1.154 (CVS log)
$Date: 2020/01/15 15:52:18 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.154  of Date: 2018/09/25 16:35:56  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00)

Default Present: Irfan, JF, paul_grenier, janina, SteveNoble
Present: Irfan JF paul_grenier janina SteveNoble
Found Scribe: Christine
Inferring ScribeNick: Christine
Found Date: 15 Jan 2020
People with action items: 

WARNING: Input appears to use implicit continuation lines.
You may need the "-implicitContinuations" option.

WARNING: IRC log location not specified!  (You can ignore this 
warning if you do not want the generated minutes to contain 
a link to the original IRC log.)

[End of scribe.perl diagnostic output]