[MPEG-OTSPEC] Questions about "cabp" feature proposal (Re: Two GSUB Proposals for OFF: 'cabp' and 'hypy')

[suzuki toshiya]

Dear Eiso,

Please let me ask a few questions on "cabp" feature.

1) The name "cabp" looks wide, general, but GB/Z 40637-2011 might have more limited context.

As the proposed description states, GB/Z 40637 is a subset of
GB 18030-2011, 14250 characters, a repertoire selected for the
typesetting of ancient Chinese books. To avoid the confusion by
the variants of the source separated or cognate characters,
GB/Z 40637 chosen a representativeb one from the cognate character
groups.

For example, "户" (sourced from GB 2312) is included, but "戶" or
"戸" (tagged GE source, introduced for the compatibility with
ISO/IEC 10646) are not included. "並" is chosen, but "竝" is not.

The characters not listed in GB/Z 40637-2011 should be untouched
by the cabp feature? If so (cabp should not change the characters
not listed in GB/Z 40637), how about the stability of the character
set of GB/Z 40637? Is there any possibility to extend it to cover
the unlisted and non-cognate characters? If there is a possibility,
the name "cabp" is slightly too generic, having some suffix to
specify the version number would be safer.

2) Is there any documentation about the glyph design policy of GB/Z 40637?

Taking a glance on the glyphs in GB/Z 40637, some glyphs have clear
difference from the G-column glyphs in ISO/IEC 10646 (like "艹" and
"++", described in the proposal - 敢, 敖, 犮, 奥/奧, 鵖 ...), but
there are many glyphs which I cannot find clear differences between
G-column glyphs in ISO/IEC 10646 versus the glyphs in GB/Z 40637,
like, 哀, 愛, 安, 岸, 暗, 凹, etc etc.

You may wonder "what is the problem with that? similar situations are
found for jp78, jp83, jp90, hojo and jp04, there are no JIS documentation
describing the glyph difference of them". It is true, but these features
were designed by Adobe, for Adobe-Japan1 glyph set. Although they are
not the parts of 14496-22, Adobe published detailed info which CID is
designed for which feature, like:
https://raw.githubusercontent.com/adobe-type-tools/Adobe-Japan1/master/aj17-kanji.txt

The developers can, but don't have to compare the JIS code charts by
themselves to select glyphs to be differentiated. I think it was very
helpful for the developers.

I hope similar list, or, a document describing the policies "how to
change GB 18030-savvy glyphs to conform GB/Z 40637-savvy". Is it
possible?

Regards,
mpsuzuki

On 2024/02/26 9:38, 陈永聪 via mpeg-otspec wrote:
> This is my first proposal for OFF. Please see http://cloud.caaph.com:10121/f/fed7bd2e3d/
> 
> I suggest adding two GSUB features. 'cabp' is used to support GB/Z 40637—2021, 'hypy' is used to support the special glyphs forms used for Hanyu Pinyin.
> 
> If you have any suggestions or feedbacks, please let me know.
> 
> Regards,
> Eiso
> 
> [c5f6]