W3C

– DRAFT –
I18N ⇔ CSS

15 October 2024

Attendees

Present
fantasai, florian, xfq
Regrets
addison
Chair
-
Scribe
xfq

Meeting minutes

florian: there are some things we can do for latin
… that we cannot do for other languages
… for CJK, Hebrew, Thai, other SEA languages, etc.
… there isn't a font metrics that corresponds to the top of the letters

fantasai: there's an official liaison statement on behalf of @@ to the opentype group at ISO
… I believe Chris sent

florian: in that case the action item on fantasai and me to do a write up is not appropriate
… because the write up has happened
… an interesting question is then waht?

https://www.mpeg.org/standards/MPEG-4/22/

w3c/i18n-actions#116

<gb> Action 116 sort out the various categories of things that get autospaced with koji (on frivoal, fantasai) due 2024-08-27

xfq: haven't looked at koji's issue in detail yet

w3c/csswg-drafts#11013

<gb> Issue 11013 [css-text] Use the Unicode East Asian Auto Spacing for `text-autospace` (by kojiishi) [Agenda+]

fantasai: maybe we should set up a work session about that

interaction between bidi and text-autospace

<fantasai> w3c/csswg-drafts#10803

<gb> Issue 10803 [css-text-4] [text-autospace] Spacing across element boundaries for BiDi content (by vitorroriz) [css-text-4] [Agenda+] [i18n-tracker] [Agenda+ i18n]

w3c/csswg-drafts#10803

fantasai: I think a lot of the cases the non-CJK text contains anything like it's a part of a string that contains anything other than @@1
… bidi text gets reordered on its own
… visual level @@2

florian: unless we imagine some different world where it's very common for people to be bilingual and fluent in Japanese and Arabic and write text in mixed languages all the time
… I feel there's a correct answer, but it's going to matter very, very rarely
… it will exist some of the time, it's just going to be in specialized publications

fantasai: those publications might decide that they just wrap all of their text in spans
… so taht they could style them in a variety of different ways correctly
… maybe add padding, but it doesn't work sometimes
… like at the end of the line
… I think the correct answer is define ti to be after reordering
… require that at least when the entire bidi string @@3
… not be sensitive to the internal reordering
… if the internals are a single phrase
… because I think that's the most common cases
… and the most clearly wrong if you get it wrong
… it should be defined as visual
… we have a prioritization of getteing ceratin cases right more important than other cases
… in terms of impl effort
… the user agent is not required to get it right in all cases, but must get it right for these cases

florian: makes sense to me

florian: koji is the one who's waiting on us mostly
… I need to look at it again

https://github.com/w3c/i18n-activity/issues?q=is%3Aissue+is%3Aopen+css+label%3Aneeds-resolution

https://github.com/w3c/csswg-drafts/issues?q=is%3Aopen+is%3Aissue+label%3A%22Agenda%2B+i18n%22

https://github.com/w3c/i18n-activity/issues?q=is%3Aissue+is%3Aopen+label%3Aneeds-resolution+label%3Awg%3Acss+

UTR for the text-emphasis skip property

fantasai: I think currently the spec relies on the general category property with a few tweaks
… are there issues with those definitions?

florian: I thought our worry was with a few tweaks
… and the fact that as Unicode adds more chars we might need more tweaks
… probably not often

florian: it would be nice if Unicode would just maintain this for us because that doesn't feel like something that is CSS specific

fantasai: I'm much more concerned about the text autospace

florian: we might want to keep it in our spec, that's fine

> rrsagent, make minutes

Minutes manually created (not a transcript), formatted by scribe.perl version 237 (Fri Oct 4 02:31:45 2024 UTC).

Diagnostics

Succeeded: s/langs/languages/

All speakers: fantasai, florian, xfq

Active on IRC: fantasai, xfq