W3C

– DRAFT –
APA Weekly Personalization Teleconference

22 February 2021

Attendees

Present
becky, CharlesL, janina, JF, LionelW, Lisa, Roy, Sharon
Regrets
-
Chair
Sharon
Scribe
Lionel, LionelW

Meeting minutes

Draft response to I18N issue 145 https://github.com/w3c/personalization-semantics/issues/145

John: We discussed multip[le values, and their seperators
… and an issue regarding conjugated terms, where a 'plus' and a 'space' are involved

Lisa: There was a discussion of the HTML reference. Issue 145.

Becky the case: a symbol, a conjugated symbol, then another symbol [symbol symbol + SPACE symbol]

JF: By listing multiple numeric values it returns a phrase
… this spacing can throw off the system

janina: hard space vs soft space?

LisaSeemanKest: We said language and directional metadata should be the default of how you parse.

<CharlesL> The above discussion is referring to Agenda 3 token lists

LisaSeemanKest: say a plus sign means conjugation (you +male, you +female)
… we are not testing translation across languages

<JF> https://w3c.github.io/personalization-semantics/content/index.html#symbol-example

JF: Example 12, <span data-symbol="13621 12324 17511">cup of Tea</span>
… The / + / is used to join the conjugated values, "her" (14707) and "name" (15691).
… <img src="her-name.png" alt="שמה" data-symbol="15691 + 14707"/>

CharlesL: All of this is in reference to Agenda item #3

JF: The space around the joiner is the cause of the problem, before and after the +

<JF> 15691 + 14707 versus 15691+14707

<LisaSeemanKest> +1

JF: The 2nd part of the question is, 'use the host language convention' but in HTML we have both space separated and comma separated

janina: We looked at ARIA to follow its lead

becky: +1 for removing the space
… (noting that parsers can handle spaces around symbols)
… ARIA references a token list and 'follow the language' with retaining the ambiguity over HTML 5

JF: Which choice do we take? (1) space (2) comma (3) ambiguous

janina: Suggest we check in with ARIA, as they have dealt with this ambiguity.

CharlesL: ARIA is 'out of the box' and published, they have to live with whatever they've got
… this is a newer specification, we can specify conjugation (without space) and go with it

JF: Should we ask TAG, WHAT-WG? Present the use case/scenario and the three choices, and get an opinion on what is the best notation.

janina: OK to go to TAG/WG

becky: Going to WHAT-WG sounds good

<JF> +1 Becky

janina: we want to avoid parser conflicts

<Zakim> JF, you wanted to propose writing up a draft question for TAG and WHAT WG

janina: send to main APA list

<JF> +1 to Janina - keeping our friends close

JF: I will write up a draft and send it to the personalization list, if everyone approves on the list then JF will send it

LisaSeemanKest: The order might be important. First APA, then WHATWG then TAG

<Zakim> JF, you wanted to ask ARIA why they chose what they chose?

janina: Big benefits from being 100% aligned with ARIA, as browsers and AT will need to parse both

JF: That's why it's worth asking why they landed on option (3)

janina: And ask, what is their experience being ambiguous on choice (3)

Sharon: taking up agenda item 4

Clean up IRC action items https://www.w3.org/WAI/APA/task-forces/personalization/track/actions/open

<LisaSeemanKest> https://github.com/w3c/personalization-semantics/wiki/actions

<LisaSeemanKest> I think u should just close it

Sharon: Seeking best way to tidy up

JF: Tidied up during the call.

LisaSeemanKest: I will close the 'prefix' as we are staying with data dash for now
… 68

Draft response to I18N issue 145 https://github.com/w3c/personalization-semantics/issues/145

Review Action items wiki - https://github.com/w3c/personalization-semantics/wiki/actions

Sharon: Lisa to draft Editors note about data- changing. Lets take a look

LisaSeemanKest: These editor's notes need to be reviewed

Sharon: I see an issue as closed on 26-Jan

New business?

LisaSeemanKest: Module 1 should point to the public explainer, not to the wiki. Currently pointing to wiki

<Roy> https://w3c.github.io/personalization-semantics/content/

<becky> https://w3c.github.io/personalization-semantics/content/

LisaSeemanKest: Look inside, lower down -- multiple references need to be validated that they are pointing to the right place

Issue 160, spoke with R

becky: Remains open, as it is tagged with Module 2/3

CharlesL: The wiki explainer-- should we put a new title on it, a sign that "This is old" ? Or rename it TAG explainer

LisaSeemanKest: The front page of the wiki has a reasonably updated list -- let's add to that

CharlesL: https://github.com/w3c/personalization-semantics/wiki/Explainer-for-Personalization-Semantics

janina: I strongly advocate that the TAG "own" this situation
… "created for W3C TAG consumption"
… we need a general explainer, and at the same time, need to produce a very specific type of defined explainer.
… TAG demands we leave a lot of stuff out

<JF> Last minute FYI: related to this TF: https://lists.w3.org/Archives/Public/w3c-wai-gl/2021JanMar/0097.html

Sharon: Need a new time for the planning call/

janina: a new recruit will join -- Matthew Atkinson from APA

np Charles, thanks for your patience

Minutes manually created (not a transcript), formatted by scribe.perl version 127 (Wed Dec 30 17:39:58 2020 UTC).

Diagnostics

Succeeded: s/TAG, WG/TAG, WHAT-WG

Maybe present: John, LisaSeemanKest