JLReq TF meeting notes 2023-8-8

[kidayasuo]

JLReq TF meeting notes 2023-8-8

(Japanese translation follows)

attendee: atsuda, atsushi, kida, kobatake, murata, nat, tahara, tajima, tlk, toshi, yamashige

WCAG 2.2 issue

WCAG 2.2 has descriptions that do not apply to Japanese or other non-Western languages, and they need to be fixed before it becomes a W3C Recommendation. As WCAG 2.2 is planned to be an ISO/IEC 40500 and subsequently JIS X 8341-3, and as the process does not allow modifications, these issues must be fixed with WCAG. We are working with the I18n WG to solve this issue. They are planning to send a comment to WCAG. Murata-san is discussing this with the DAISY consortium. It was suggested that JLReq TF send a letter to AC Reps in Japan so they can send feedback.

Todo: Murata and Kida are to have a separate meeting to discuss the details of the action. (done)

Updates from CITPC

In the last few months, we noticed discrepancies between how CFF-based OpenType Japanese fonts have been implemented for these 20+ years and what OpenType specification says. Two examples so far are kern and VORG. Kida started a team in CITPC to review all tables in C/J fonts to find out if there are any similar issues. The goal is to propose modifications to the Open Font Format team in ISO/IEC.

We want some more engineers from font developers to participate; it would be great if someone is familiar with TrueType flavored fonts and knowledgeable about Chinese fonts.

Todo: Kobatake-san from DynaComware will invite engineers from within the company. (done. Thank you!)

Updates from Open Font Format group *1

In the past, the Open Font Format has just been incorporating changes from Microsoft’s OpenType specification. For the first time, new changes are being added to the Open Font Format side by accepting the proposals from Google.

Among them, the proposal to allow more than 64k glyphs is important to Japan, as the government’s character set to print names of people and places is around 68k. Kida was recently appointed as a member of the Japanese mirror of SC29 WG3 via CITPC. He will provide updates at CITPC’s Teams and here at JLReq TF.

*1: A group within “ISO/IEC JTC1 SC 29/WG 3 MPEG Systems group” and publishes ISO/IEC 14496-22 Open Font Format standard. The latest edition is 4th and was published in 2019. They are now preparing the 5th edition.

jlreq-d: discussion on chapter 8 (tentative numbering): how digital text differs from print

The discussion is made on and immediately reflected to the jlreq-d wiki.

Todo: kida will draft text based on the items discussed. We will review it at the next JLReq TF meeting.

The next meeting will be on 8/29 Tuesday 10 am.

---

（日本語。chatGPT による自動翻訳）

WCAG 2.2 問題

WCAG 2.2には、日本語や他の非西洋言語に適用されない内容が含まれており、これがW3Cの勧告になる前に修正する必要があります。WCAG 2.2はISO/IEC 40500およびその後のJIS X 8341-3として計画されており、プロセス上修正が許されていません。よってこれらの問題はWCAGで解決しなければなりません。この問題を解決するために、I18n WGと協力しています。WCAGにコメントを送る予定です。Murataさんがこの件でDAISYコンソーシアムと協議中です。JLReq TFが日本のAC Repsに手紙を送ることが提案されました。

Todo：MurataとKidaがこのアクションの詳細について別途会議を持つ（完了）

CITPCからのアップデート

過去数か月で、CFFベースのOpenType日本語フォントが20年以上にわたってどのように実装されているかと、OpenTypeの仕様が何を言っているのかに違いがあることに気づきました。これまでのところ、kernとVORGの2例があります。KidaはCITPC内で全てのC/Jフォントのテーブルをレビューし、同様の問題がないか調査するチームを開始しました。目標は、ISO/IECのOpen Font Formatチームに修正を提案することです。

TrueTypeフレーバーのフォントや中国のフォントに精通したフォント開発者からもっと多くのエンジニアが参加することが望ましいです。

Todo：DynaComwareのKobatakeさんが、社内のエンジニアを招待する（完了。ありがとうございます！）

Open Font Formatグループからの更新情報 *1

過去には、Open Font FormatはMicrosoftのOpenType仕様からの変更を単に組み込んでいましたが、初めてGoogleの提案を受け入れて、Open Font Format側に新しい変更が加えられています。

これには、64k以上のグリフを許可する提案が含まれており、これは日本政府の人名や地名を印刷するための文字セットが約68kであるため、日本にとって重要です。Kidaは最近、CITPC経由でSC29 WG3の日本のミラーグループのメンバーに任命されました。彼はCITPCのTeamsおよびこちらのJLReq TFで更新情報を提供します。

*1：「ISO/IEC JTC1 SC 29/WG 3 MPEG Systems group」内のグループであり、ISO/IEC 14496-22 Open Font Format標準を公開しています。最新版は第4版で、2019年に出版されました。現在、第5版を準備中です。

jlreq-d：第8章（仮番号）についての議論：デジタルテキストが印刷とどう違うか

議論は、即座にjlreq-dのwikiに反映されています。

Todo：kidaが議論された項目に基づいてテキストを作成する予定です。次のJLReq TF会議でそれをレビューします。

次回の会議は8/29（火）10時です。

[xfq]

In the last few months, we noticed discrepancies between how CFF-based OpenType Japanese fonts have been implemented for these 20+ years and what OpenType specification says. Two examples so far are kern and VORG. Kida started a team in CITPC to review all tables in C/J fonts to find out if there are any similar issues. The goal is to propose modifications to the Open Font Format team in ISO/IEC.

We want some more engineers from font developers to participate; it would be great if someone is familiar with TrueType flavored fonts and knowledgeable about Chinese fonts.

If there is a detailed description of use cases and font issues/gaps, I'd be happy to ask Chinese font developers.

[kidayasuo]

Hello Fuqiao, That is super. There would be a few options of doing the investigation depending on the amount of time they wan to use. a. The team at CITPC identify discrepancies, and font developers in China verify if they are also true for fonts they make. b. Setup a semi-joint investigation. Each conducts the investigation and the result will be shared through a periodic joint meeting. Producing single proposal. and anything between a and b. So far issues are found with CFF flavored OpenType fonts. - kida

…

2023/08/25 10:34、Fuqiao Xue ***@***.***>のメール: In the last few months, we noticed discrepancies between how CFF-based OpenType Japanese fonts have been implemented for these 20+ years and what OpenType specification says. Two examples so far are kern and VORG. Kida started a team in CITPC <https://moji.or.jp/> to review all tables in C/J fonts to find out if there are any similar issues. The goal is to propose modifications to the Open Font Format team in ISO/IEC. We want some more engineers from font developers to participate; it would be great if someone is familiar with TrueType flavored fonts and knowledgeable about Chinese fonts. If there is a detailed description of use cases and font issues/gaps, I'd be happy to ask Chinese font developers. — Reply to this email directly, view it on GitHub <#373 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AH5C6A4VJPTDGE4GXWD6DZDXW76LFANCNFSM6AAAAAA35XNTRE>. You are receiving this because you authored the thread.