W3C

– DRAFT –
Adapt Working Group Weekly Teleconference Monday May 1st 2023

01 May 2023

Attendees

Present
janina, Lionel_Wolberger, matatk, mike_beganyi, Russell, Sharon
Regrets
-
Chair
-
Scribe
janina, matatk

Meeting minutes

A11y4Children follow-ups

Issue 240 update, "Could we build symbolic annotations with existing Web standards?"

A11y4Children follow-ups

Lionel_Wolberger: Notes the document from Web accessibility for Children Community Group

https://docs.google.com/document/d/18CDOklgwNYK2gILIz_IADeU_TiyQW6kuJWfP3Nhfk1c/edit?usp=sharing

Lionel_Wolberger: Finding well formed opinions based on years of experience

Lionel_Wolberger: They address areas I believe we haven't

Lionel_Wolberger: proposing new attributes, etc

Lionel_Wolberger: we'll need to consider how to work with this

matatk: Didn't see it via our archive

matatk: Believe some overlap to COGA, e.g. trauma

matatk: phps our concern is to be careful of how many attributes and not overdo

matatk: think the what rather than the how may be a useful approach?

janina: I've also not read it yet, but +1 to matatk

Lionel_Wolberger: Notes their Google Doc is packed with compelling ideas

Lionel_Wolberger: we'll have to schedule processing this

Lionel_Wolberger: asking what our next steps should be?

matatk: agree we need a bit of time to process this

matatk: Do we have existing expertise -- is there sufficient staffing to work on it?

matatk: Should note finishing symbols is a priority, but that also gives us a bit of time

matatk: suggests pinging COGA for staffing

matatk: we need to do some expectation management

mike_beganyi: agree on collab review and synthesizing priorities

janina: Fine with us consulting COGA on guidance, use cases, etc. but they may not be able to help with the technological/architectural issues.
… With respect to UDL, I think the question will be: how few tags can we introduce to meet most of the use cases?
… Are they generalizable in such a way that we can re-organize them on the fly? E.g. in some cases, a H1 may become an H2. I think EPub/DPub would be interested in this (from the conversations we've had).
… RQTF is happy to work on this [UDL] but they need specific questions from us.

Lionel_Wolberger: re-organizing?

janina: You might have a recipe book; the H1s could be e.g. Chicken, or they could be Soups.
… Is that where the CG's thinking is?

Lionel_Wolberger: Will return to this hopefully with some ideas of what next

matatk: agrees

Issue 240 update, "Could we build symbolic annotations with existing Web standards?"

Lionel_Wolberger: issue 240 with over 3K words

Russell: has this grown? Seems there's more there now

Lionel_Wolberger: Perhaps; latest is last week

Russell: I know him

Russell: Doing unicode for BCI; the registrar for the subtag registry in the unicode group

Russell: proposing rather than the BCI ids we use the unicode encoding

Russell: I agree it will be widely used once established

Russell: concern is that it would be using word spellings for ids, not sure that's what we want

Russell: believe we would want unique ids

Russell: notes that it would be 0x.. + 0x.. something else rather than our ids

russel also bci can change; they're generative

Russell: over time what once had meaning ceases to have meaning

Russell: the unique id wouldn't change, though its representation could

russel somewhat like the same word, different spellingrusselonce unicode is there it wouldn't change, only expanded

Russell: we do have ids for each unicode; but also id for concatinated unicode chars

Russell: puts me still on the fence here

Russell: bci codes are dependent on bci which should persist, but phps an unnecessary dependency

Russell: someone will need to maintain the db

Russell: the symbols aren't self explanatory, so concepts need to be maintained

Lionel_Wolberger: we need a primary key, and the unicode may not quite function that way

matatk: very helpful discussion

matatk: not up to date on thread but 3 concerns

matatk: whether unique values are the symbols or the building blocks for the symbols

matatk: understood both

matatk: I've seen symbols get deprecated, changed, etc

matatk: was that accurate?

matatk: we need to know whether we could manage a concatinated set of 0X values

Russell: correct

Russell: some of this is historical to bliss and how contractions, etc, work

The previous standard Russell mentioned: https://en.wikipedia.org/wiki/ISO-IR-169

Russell: notes first was pictographic but then parsed for component parts most of which (but not all) are meaningful on their own -- apparently about 1500

Russell: unicode cannot fully display every bliss construction

Russell: as unicode evolves, though, it will likely do even better at supporting bliss

Russell: the unicode activity is impacting bliss extensions; people consider it as they develop symbol construct

Lionel_Wolberger: let's see what our next action might be on this ...

Lionel_Wolberger: shares the registry screen

Lionel_Wolberger: if we understand 240 as a suggestion to change our primary key

Lionel_Wolberger: if it's pretty much a straight up replacement of ids for unicode values or concatinated unicode values that could be ok

Lionel_Wolberger: looking at an example with "mind"

Russell: great example

<Lionel_Wolberger> 24488 drug,mind-altering_drug

Russell: "mind altering drugs" would consist of two values

<Lionel_Wolberger> 15471 mind,intellect,reason

Russell: have actually done it that way, as a compound

<Lionel_Wolberger> 20518 sport

Russell: similarly "mind sports"

Lionel_Wolberger: thinking the ambiguity and ordering makes unicode a less useful key

Russell: certainly less readable

<Lionel_Wolberger> Proposed response: The topic was discussed at [put offset]

<Lionel_Wolberger> ... we do not find that the unicode spellings will substitute for the ID as we currently use it

<Lionel_Wolberger> ... which constitutes a nonambiguous, unique "Primary Key"

<Lionel_Wolberger> ... in the sense of a Primary Key in SQL and relational database theory

<Lionel_Wolberger> 21799 mind_sports

matatk: aren't we already sometimes concatinating?

Russell: sometimes when the same concept is repeated --

<Lionel_Wolberger> ... we are discussing it, we are not sure that this proposal will substitute for the 'primary key'

Minutes manually created (not a transcript), formatted by scribe.perl version 210 (Wed Jan 11 19:21:32 2023 UTC).

Diagnostics

Succeeded: s/where their/where the CG's/

No scribenick or scribe found. Guessed: janina

All speakers: janina, Lionel_Wolberger, matatk, mike_beganyi, Russell

Active on IRC: janina, Lionel_Wolberger, matatk, mike_beganyi, Russell, Sharon