Richard, fantasai, and Florian did not object. We will update the document by merging pending modifications as a draft note and make it a discontinued draft afterward.
Todo: kida to send mail to i18n to officially communicate the decision. (not done as of 8/25)
How to reduce necessary computation to fold ruby
Kida discussed this with Nat. He believes it will significantly simplify by disabling ruby overhang. We could reduce the impact by disabling the overhang only when the ruby is folded. Or, as the effect of the overhang is more prominent with shorter rubies, we could also disallow folding them.
Nat also wanted jlreq-d to describe sophisticated methods along with the simplified method. Agreed. However, we want to provide a way to simplify ruby; at the same time, we want to discuss issues with the simple method and invite people to consider implementing it more sophisticatedly.
Todo: Kida to discuss Ishii-san go get his feedback (not done as of 8/25)
Discussion re: jlreq-d layout model
Kobayashi-san drafted a document comparing the layout model between JLReq (print) and JLReq-d (digital), and we did free-form discussions related to the model. The Kobayashi-san’s text was copied here.
Takeaways I could capture from the discussion are:
The layout on print is based on fixed-size pages, and the one on digital devices is based on dynamic, reflowable scroll.
JLreq-d should focus on the latter model and leave the paged model to the original JLReq (agreed).
Adjusting the amount of space around the text area is considered important in Japanese text layout. (Nat. I needed clarification on exactly why it is specific to Japanese text.)
The line length needs to be multiple of fullwidth for Japanese text with full justification (which is the default) because otherwise, space needs to be inserted between each character.
Factors that make text look beautiful and factors that are necessary (in terms of what?) are two different things. We need to discuss them separately.
We need to be careful about discarding old practices because oftentimes, things regarded as obsolete still hold value.
The most important thing is to expand options/capabilities. We would then provide information in jlreq-d so readers can make informed decisions.
Readers typically do not complain even if the layout is hard to read because you can read it. Ones who create, however, should strive to provide a better layout.
Web designers design layouts of areas but often just throw text into boxes and do not care about how they are laid out.
Kida proposed developing a draft TOC, the first section first. It will exhibit his vision. After reviewing the TOC and the first section, we can start distributing roles.
Todo: Kida to develop a draft TOC and the first section by the next meeting (done).
JLReq TF meeting notes 2023-6-20
（日本語 follows）
Attendees: atsushi, kida, murata, nat, shinyu, suzuki, tajima, tlk, toshi
Discontinuing simple-ruby document
Richard, fantasai, and Florian did not object. We will update the document by merging pending modifications as a draft note and make it a discontinued draft afterward.
Todo: kida to send mail to i18n to officially communicate the decision. (not done as of 8/25)
How to reduce necessary computation to fold ruby
Kida discussed this with Nat. He believes it will significantly simplify by disabling ruby overhang. We could reduce the impact by disabling the overhang only when the ruby is folded. Or, as the effect of the overhang is more prominent with shorter rubies, we could also disallow folding them.
Nat also wanted jlreq-d to describe sophisticated methods along with the simplified method. Agreed. However, we want to provide a way to simplify ruby; at the same time, we want to discuss issues with the simple method and invite people to consider implementing it more sophisticatedly.
Todo: Kida to discuss Ishii-san go get his feedback (not done as of 8/25)
Discussion re: jlreq-d layout model
Kobayashi-san drafted a document comparing the layout model between JLReq (print) and JLReq-d (digital), and we did free-form discussions related to the model. The Kobayashi-san’s text was copied here.
Takeaways I could capture from the discussion are:
The row conversation is recorded below (which does not necessarily capture everything). https://www.w3.org/2023/06/20-jlreq-minutes.html
How do we develop jlreq-d?
Kida proposed developing a draft TOC, the first section first. It will exhibit his vision. After reviewing the TOC and the first section, we can start distributing roles.
Todo: Kida to develop a draft TOC and the first section by the next meeting (done).
the next meeting will be on 7/11 10 am JST.
simple-ruby document の廃止について
Richard、fantasai、およびFlorianは反対意見を示さなかった。保留中の修正をマージしてドラフトノートとして文書を更新した後、廃止される予定です。
Todo: Kidaがi18nに正式に決定を伝えるメールを送る（8/25現在、未実施）。
ルビを折りたたむために必要な計算を減らす方法
KidaがNatとこの問題について議論した。ルビのオーバーハングを無効にすることで、大幅に簡略化できると考えています。ルビが折りたたまれる場合にのみオーバーハングを無効にすることで、影響を軽減できるかもしれません。また、オーバーハングの影響が短いルビでより顕著であるため、それらを折りたたむことを禁止することもできます。
Natも、簡略化された方法と共に、より高度な方法をjlreq-dで説明するべきだと考えました。同意。ルビを簡略化する方法を提供する一方で、簡略化方法の問題点について議論し、より高度な実装を検討するように人々を招待したいと考えています。
Todo: KidaがIshiiさんとフィードバックを得るために議論する（8/25現在、未実施）。
jlreq-dのレイアウトモデルに関する議論
Kobayashiさんが、JLReq（印刷）とJLReq-d（デジタル）のレイアウトモデルを比較する文書を作成し、関連する自由形式の議論を行いました。Kobayashiさんのテキストはこちらにコピーされています。
議論から得られた主要なポイントは以下の通りです：
Todo: Kidaが次回の会議までにドラフトの目次と最初のセクションを作成する（完了）。
次回の会議は7/11 10時JSTに開催されます。