W3C

– DRAFT –
Personalization Task Force

15 Mar 21

Attendees

Present
becky, janina
Regrets
-
Chair
Lisa
Scribe
becky

Meeting minutes

<janina> Meeting Personalization Task Force

<JF> trackbot, please join

<JF> s/

Draft response to I18N issue 144 & 145 (John)

<LisaSeemanKest_> https://github.com/w3c/personalization-semantics/issues/145

LisaSeemanKest_: 2 issues from i18n

jf: have drafted response to 145

JF: #145 has concerns about us republishing the data purpose values from WCAG 2.1 and HTML5 spec. That list of terms is hosted and maintained by WHAT-WG. We want to manage the list so that changes don't break our spec.

JF: our TF supports the creation of a W3C repository for these types of glossaries and terms, have brought that up to APA

JF: drafted a response to I18N that we are pursuing with WAI leaders via APA to host glossaries

JF: regarding calendar concerns we need to make a change to specify the format or how to determine it - needs to be discussed withn the group

<JF> https://html.spec.whatwg.org/multipage/form-control-infrastructure.html#attr-fe-autocomplete-language

LisaSeemanKest_: COGA identifies that must specify correct language codes, is that enough?

JF: WHAT-WG spec does reference language code and specifies the value type and specific type; We should update our code to specify to use BCP or ISO language code

<JF> +1 to janina

janina: We need to be precise about what "correct language code" means in our spec. Believe that I18N should be providing guidance on the correct type to use.

<JF> https://en.wikipedia.org/wiki/List_of_ISO_639-1_codes

ack, janina

<Matthew_Atkinson> https://html.spec.whatwg.org/multipage/dom.html#the-lang-and-xml:lang-attributes

Janina: this is important for in context lang attributes

<JF> https://www.w3.org/International/questions/qa-html-language-declarations

<LisaSeemanKest_> https://raw.githack.com/w3c/coga/consistency_checks/content-usable/index.html#more-details-14

LisaSeemanKest_: COGA references ISO and BCP language codes

<JF> https://www.rfc-editor.org/rfc/bcp/bcp47.txt

JF: BCP 47 is from IETF

JF: BCP 47 references ISO

Janina: we need to make sure whatever we select in w3C includes sign language

<janina> https://www.evertype.com/standards/iso639/sgn.html

<JF> "ISO 639-1:2002, Codes for the representation of names of languages -- Part 1: Alpha-2 code"

<JF> language Ability to select the preferred language for the content.

LisaSeemanKest_: I remember adding a note that should use local data codes, we should verify that note exists and update to make more specific with help from I18N

JF: we currently don't specify how to declare the value

<LisaSeemanKest_> Use language tags. Language tags are the key means to achieve the goal of unambiguous text formatting. See HTML [HTML] language tags and BCP 47 Language Codes.

<JF> <input type="text" name="12345" data-purpose="language"/>

LisaSeemanKest_: Think we need a sentence similar to what is in COGA about language and ask I18N what specifics they want

JF: the way we are using language is as an attribute value, the purpose of the input is to ask the user what language they prefer.

<JF> <label for="foo">What is your prefered language</label> <input type="text" name="12345" id="foo" data-purpose="language"/>

JF: I am just collecting their language

janina: but many people wouldn't know what to use

<LisaSeemanKest_> ageda+ editors nots need to be updated and removed

JF: but we are saying the purpose of the input is to collect the user's preferred language

JF: I think I18N is thinking about output to the screen for human reading but we are talking about machine readable content

JF: this input example is just collecting data - date about the pereerred language

Janina: what are we doing we the data?

LisaSeemanKest_: User agents need to understand how to convert to dates etc. to a format that the user understands

<JF> +1 to Matthew re: validation

Matthew_Atkinson: there are 2 separate things going on, we are trying to establish the purpose of this field - there are two steps, identify the purpose and validation that the selection/input is valid. Whose responsibility is to validate?

<JF> <form lang="de"> <label for="foo">Was ist Ihre bevorzugte Sprache?</label> <input type="text" name="12345" id="foo" data-purpose="language"/> <submit> </form>

JF: John believes is it the User agent/AT responsibility

JF: purpose of the input is just to capture the preference, the lang attribute on the page specifies the currently displayed language

<JF> data-purpose="cc-number"

Janina: but need to make certain that we communicate in the correct language

JF: purpose of the attribute is to tell machines what type of information we are capturing, name, address, cc-number, etc. language is just another of those values

LisaSeemanKest_: doesn't think that is the only use case

<JF> The purpose attribute provides the context of a text input field such as a text box. It is typically used on an input of type text, or element with a corresponding role.

LisaSeemanKest_: AT could also use the cc-number info to display a particular icon, or include a warning that the user is going to disclose the cc-number;

JF: yes, I agree, but you could do the same thing with language - display an icon that represents a language collection. This is just about identifying the purpose of the input

<JF> Defining context is not an internationalization issue, it is **MACHINE READABLE TEXT**

<JF> <html lang="fr">

LisaSeemanKest_: looking at date - perhaps display a specific date picker with icons for the various days; as an AT developer I want to know what date format is desired, so would be useful to use the date semantic to help with data binding

JF: the point of our attribute is machine readable text

LisaSeemanKest_: I just want to add a note that page should also be using lang attribute correctly

<JF> +1 BEcky

becky: agree we are just providing machine readable info to help with selecting the correct user preference and matching to language code of the page

JF: I believe I18n is not understanding that the language value in this case is just machine readable

LisaSeemanKest_: suggest drafting a note that language must be specified on the page and asking I18N if that is sufficient

JF: this particular attribute is not generally used for output on the screen in most cases, we are just saying what the purpose of the input is - it is to collect the language

LisaSeemanKest_: Do we just want to explain JF's point to I18N or to add a sentence about, Note we recommend author's should be aware that the language standard should be used.

<LisaSeemanKest_> note ather should be awear that the lang atandard , and should also be used

Do people want to add such a note?

<LisaSeemanKest_> add the sentence ?

<LisaSeemanKest_> +1

-1

<JF> -1

<janina> +0

<JF> "Please note that 'language' is an attribute VALUE, and not an attribute seeking which language is being used. Our attribute (@purpose) is intended to be machine-readable text. Content authors will also need to ensure that language and bi-directional concerns *related to the form* are addressed, but this attribute is not the vehicle for these concerns"

<LisaSeemanKest_> we recomend useing a valid bcb 427 code and valid lange tags inside the code

Matthew_Atkinson: believe the main issue is addressing the confusion by I18N but not sure this is the right way to do it. Will this go away when we have a defined glossary. If I18n is confused will others be?

LisaSeemanKest_: I am hearing if we get a correctly drafted sentence, most people are willing to include

JF: clarifying the language is an attribute value, it is not an actual value

<LisaSeemanKest_> Please note that 'language' is an attribute VALUE, and not an attribute seeking which language is being used. Our attribute (@purpose) is intended to be machine-readable text. Content authors will also need to ensure that language and bi-directional concerns *related to the form* are addressed, but

JF: see proposed text above

LisaSeemanKest_: I think this should go at the top rather than just for language

Janina: it is not just language it is more locale related;

LisaSeemanKest_: asks Janina to wordsmith the proposal from John

<Matthew_Atkinson> +1 on the clarification about these being values (not attributes), in the way JF mentioned, and +1 to LisaSeemanKest_'s suggestion of making this a generic note closer to the top.

Janina: if i18n is confused others may be as well

JF: this attribute only applies to form inputs

janina: not sure we agree

<Zakim> JF, you wanted to note that @purpose is applicable to form inputs ONLY

a bit of back and forth about the use of the purpose attribute

<LisaSeemanKest_> -1 to john , i dissagee

<JF> YES

<JF> The purpose attribute provides the context of a text input field such as a text box

<LisaSeemanKest_> action, janina, to write wording on please use standards for dates etc

Action: janina, to write wording on please use standards for dates etc

becky: maybe we are not using it for forms, but WCAG specifically says it is for forms. even if we are using it for other purposes, to display an icon, then that piece of data - language preference - needs to be stored

still disagreements within group

Summary of action items

  1. janina, to write wording on please use standards for dates etc
Minutes manually created (not a transcript), formatted by scribe.perl version 127 (Wed Dec 30 17:39:58 2020 UTC).

Diagnostics

Succeeded: s/trackbot, please join/

Maybe present: jf, LisaSeemanKest_, Matthew_Atkinson