Meeting minutes
<addison> trackbot, prepare teleconference
new member
Fatima is introducing herself to the group
fatima: hope to work e.g. on translating material
richard: you do not have the technical background we have ...
… but there are many areas in which you can make a great difference
fatima: happy to help
Agenda Review
Action Items
<addison> https://
action-1034?
<trackbot> action-1034: Addison Phillips to Triage awaiting comment resolution items -- due 2021-06-24 -- OPEN
addison: will work on this during this call and others, keep this open
action-1039?
<trackbot> action-1039: Addison Phillips to Propose pr for https://github.com/whatwg/html/issues/4814 -- due 2021-07-01 -- OPEN
addison: I made the changes after our discussion
… Richard made further comments, these are now merged in
close action-1039
<trackbot> Closed action-1039.
action-1048?
<trackbot> action-1048: Addison Phillips to Create PR on webauthn with new metadata proposal -- due 2021-07-21 -- OPEN
addison: topic for our call
… keep the action open
action-1051?
<trackbot> action-1051: Addison Phillips to File issues about domexception against webidl to add metadata and to provide for localizability -- due 2021-07-22 -- OPEN
addison: pending
action-1061?
<trackbot> action-1061: Richard Ishida to Remove encoding tests and put a link to wpt tests -- due 2021-08-26 -- OPEN
richard: done
close action-1061
<trackbot> Closed action-1061.
action-1062?
<trackbot> action-1062: Richard Ishida to Review open css-text-decor issues for close or escalation -- due 2021-08-26 -- OPEN
richard: ongoing
action-1063?
<trackbot> action-1063: Addison Phillips to Respond to vocab-dcat issues pinging for updates -- due 2021-08-26 -- OPEN
ongoing
action-1064?
<trackbot> action-1064: Felix Sasaki to Propose health warning text on vocab-dcat and language -- due 2021-08-26 -- OPEN
waiting for discussion on language tag reg. discussion
action-1067?
<trackbot> action-1067: Addison Phillips to Follow up on previous action related to language tag registry online -- due 2021-09-02 -- OPEN
ongoing
action-1066?
<trackbot> action-1066: Richard Ishida to Move pronunciation-gap-analysis issues to new document -- due 2021-08-26 -- OPEN
ongoing
action-1070?
<trackbot> action-1070: Addison Phillips to Reply to css about scrollbars -- due 2021-09-09 -- OPEN
pending
action-1071?
<trackbot> action-1071: Richard Ishida to Send uax50 change request document for review by wg -- due 2021-09-09 -- OPEN
done
close action-1071
<trackbot> Closed action-1071.
action-1072?
<trackbot> action-1072: Fuqiao Xue to Send feedback from clreq to jlreq on uax50 proposed changes -- due 2021-09-09 -- OPEN
done
close action-1072
<trackbot> Closed action-1072.
action-1073?
<trackbot> action-1073: Atsushi Shimono to Coordinate citpc and w3c publication of uax50 change request -- due 2021-09-09 -- OPEN
ongoing
action-1074?
<trackbot> action-1074: Addison Phillips to Respond to personalization-semantics that we can meet the 13th at 1400z -- due 2021-09-09 -- OPEN
done
close action-1074
<trackbot> Closed action-1074.
action-1075?
<trackbot> action-1075: Addison Phillips to Respond to css confirm 27 oct at 1600z for tpac crossover meeting -- due 2021-09-09 -- OPEN
done
close action-1075
<trackbot> Closed action-1075.
action-1076?
<trackbot> action-1076: Fuqiao Xue to Review css open issues for important items to resolve in tpac -- due 2021-09-09 -- OPEN
done
close action-1076
<trackbot> Closed action-1076.
Info Share
richard: next week is geek week
… I will probably be at our meeting, but will not have done much prep
RADAR Review
<addison> https://
addison: two new requests
… richard looked at CSS display 3, did not find anything
addison: ok to move the spec to done?
richard: fine if others are OK with that
… it is in CR at the moment
… there have been some additions, I did not see any issues
addison: so move this to completed?
Fine by me
richard: ok
addison: ok
addison: moving to "secure payment confirmation"
Action: addison: ping ian about joint meeting during tpac in re: secure payment confirmation
<trackbot> Created ACTION-1077 - Ping ian about joint meeting during tpac in re: secure payment confirmation [on Addison Phillips - due 2021-09-16].
richard: they have not done their own self-review
addison: will mention that
UAX50 change request
richard: have people read the request, do you have any requests?
addison: xfq wrote a mail about chinese ... I did not see controversial things
richard: can we vote to move forward as a group?
addison: so that the doc has the group endorsement
<r12a> +1
<atsushi> +1
<xfq> +1
<David> +1
<addison> +1
+0
approved
Resolution: working group approved publication of UAX50 request
atsushi: will bring proposal from current wiki to docs
… will prepare other things and bring them back to the WG
… without touching the main content
request to close payment-request issues
<addison> https://
<addison> https://
richard: I did not look into this in detail, also did not do the review
… a bunch of the comments is like: they removed the features from the spec
… so we can close these issues
… in other cases they say: we do this not in this but in the next version
… that is worrying since you do not know if there will be a next version
… we should put a paragraph in the spec saying: these are issues we have discussed with the i18n wg, and will consider them in the next version
… for things which are not in the spec
… we want to make implementers aware that there may be additonal changes
Action: addison: review payment-request issues
addison: will take an action to review this
<trackbot> Created ACTION-1078 - Review payment-request issues [on Addison Phillips - due 2021-09-16].
addison: is it OK to close issues we were satisfied by in the past?
… as you say, Richard, we can insert health warnings if they do not intend to address things
+1
addison: I will close the related issue, as part of my action item
<addison> https://
<addison> https://
<addison> https://
addison: in a webauth spec, the web auth people invented an approach to indicate language and directionality
… also sent a note to Martin about this, did not yet see a response from him
… above links show my proposed changes to the spec
… I did have a conversation with the group a few weeks ago, reported on that
… they prefer to append information at the end
… I ignored the "terminating character (sequence)" piece
david: the terminating character gives them a signal
addison: correct, they wanted that to indicate that the string has been truncated
… using non ascii characters increases the byte count
… the encoding I proposed is inspired by JSON-LD, but not the same
… risk of using characters that are more common: more likelyhood to be confused with real strings
richard: we may set a precedent here
… we are designing a way to do what people should not do
… and are coming up with a new way to do that
… also not sure if it is our job to solve the problem
… we should give them advice, e.g. like "use the ascii character" or "you may need a final character" and "look at json-ld"
… but rather than writing their text, we should say: go and research this how to do it
addison: we did that. I tend to agree with Richard
… maybe I should insert a health warning in their text
… it does not say why the string serialization exists
… if we had a big red box saying: "this was invented for compatibility with a wide array of low capability devices, do not copy this in your spec"
richard: we are inventing a tech solution, without research and review
discussion on writing text vs. telling how to write the text
richard: they need to dicuss how to do this
addison: I could say: these are the decisions you need to make
… how to encode language tags, direction etc. ...
richard: that would be better
addison: I would have this PR to edit it, with a box: you need to decide how to handle language and direction metadata
richard: and put the advice in the issue
addison: will do
… what about the details I put?
… about language tags I said: BCP47, folded to lower case
"If language metadata is present it is encoded as a sequence of ASCII code points representing a language tag mapped to lowercase. "
richard: this is like writing the spec
addison: nobody will disagree with using ASCII code points
richard: in charmod, we made a distinction between case folding and case mapping
addison: refrained from linking to charmod
richard: just worried about "mapped" to lower case
addison: understand
addison: about directionality, they had rlm and lrm
… with my proposal the ascii sequences can be truncated, but still will be unambigous
… most of the strings will never be truncated
… and the string will remain unambigous even if truncated
richard: you do not want to see ltr and rtl on screen
… if you had underscore followed by lrm or rlm, it should be removed before display
addison: naive displays could not handle this
addison: would we recommend a different mechanism than ascii strings for direction?
no suggestions for other mechanisms
richard: if you display these things they are bidirectional
addison: agree
david: on a naive device, that may create issues
addison: the language tag would display OK
… but things could be displayed on the wrong side
richard explains further on the directional run in the example
<r12a> حبیبالرحمان^^^ar-sa_rtl
richard: in the browser this is a left-to-right directional context
addison showing a demo of bidi issues
david: had similar issues decades ago in pager systems, seem to come back
addison: does that mean anything for what we would suggest?
richard: do not think so
addison: I will remove our recommendation on how to decorate the string
… and keep the language related recommendation?
richard: I would remove all concrete suggestions
… the two paragraphs plus the bullet list
… in the issue you can have an example of how to do things
addison: that is what I intend to do
… if we decide not to do the pull request
… I would say in the issue: here is where you need to make decisions
… meta question is: do we need to make recommendations?
… saying: here is the list of choices, and here is our recommendation
richard: this will take a long time to come up with
… these guys will better know what fits into the industry
addison: they are guessing, and they want us to tell them how to do this
… should I prepare a proposed comment until next week?
all agree
addison: we should also put these things into string meta, although that won't be in time for this group
… we could write down the considerations in string meta
… my concern is: see e.g. json-ld
… I know json-ld exists
… having more ways to do this is not the right things
david: could it happen that they put other metadata into the string?
addison: we do not want this
<r12a> https://
richard: string-meta is the place to start to write about this
<FatimaBana> (thank you)
richard: describe situation carefully in which you do not follow the advice we give in string meta
… we need to say: you need to have very good reasons for doing this
… and we need to describe those reasons
Action: addison to prepare comment for web auth until next week
<trackbot> Created ACTION-1079 - Prepare comment for web auth until next week [on Addison Phillips - due 2021-09-16].
addison: AOB for this week?
nothing, adjourned