一切经音义三种校本合刊

一切經音義 朝鮮高宗二十三年至三十八年（1236-1251）大藏都監刻本

一切經音義 日本大正昭和間（1926-1931）東京大正一切經刊會鉛印暨影印本

As mentioned in the meeting by Wang Yifan, the same character was submitted in WS2017 as 03913 and was postponed because the text suggests it is a miswritten ligature.

Keep it as a stable miswritten form.

Yes, this form is stable and not error, but the main and important issue is if it is suitable to encode an obvious ligature as a separate CJKUI. If yes, maybe I can provide hundreds even thousands of this kind of characters in future; if not, we can try to use other method to handle it by GSUB.

This is a typesetting issue beyond IRG.

On the other hand, this is a ligature of 蟲 insect and 鳥 bird, it is not easy to confirm which radical is suitable, both are right, but both are wrong.

This is page to show a demo to use 'liga' and 'ccmp' to handle the CJK ligature with WOFF. Note: this website will be removed when we finished the discussion.
http://ccamc.org/WS2021-liga/liga.xml



If you are interested in the font, please find here. https://gitee.com/eisoch/cjkliga

The 'liga' solution on CJKUI has been used by FounderType (方正字库) in 可口可乐在乎体. If any IRG expert thought this solution is not good, please also let FounderType know. Please see Section of the following page. Note: The term “EUDC” is not suitable to use here, and we all know they mean PUA, but there is no need to use PUA for this solution.
https://mp.weixin.qq.com/s/3A49hplUiiL5iecJJeJwig

'ccmp' solution is the better one for me, and I add zwj between two elements.

If any IRG expert hasn’t known about OFF/OpenType and WOFF, please read ISO/IEC 14496-22:2019 and it Amd.1:2020, and the related W3C recommendation first.

Strongly support not encoding as cjkui simple ligatures of two characters (i.e. where two characters are simply squeezed together into a single character space for aesthetic or practical purposes) where there is no semantic difference between the ligatured and unligatured forms.