source-han-sans icon indicating copy to clipboard operation
source-han-sans copied to clipboard

Regarding the 尔 and 欠 (dis)unification

Open Marcus98T opened this issue 1 year ago • 3 comments

UPDATE: Please see this comment for my latest sentiment. The original post was a weak argument to try and force certain region shapes on others when I have not considered that other people may find it uncomfortable to look at.


I apologize if I didn't post in my own issue thread about glyph consolidation because I think nobody would be aware of any new posts in an old long thread.

It is mentioned before that the current CN/TW/HK version of the components of 尔 need to be dis-unified because apparently the 小 part and the top part cannot connect as part of the 新字形 rule. ~That will unfortunately add more what I believe to be unnecessary glyphs and I rather have the JP and KR disconnect the two parts in 尔 if the glyph modification of the 尔 component is planned to go ahead (like in Source Han Serif). In other words the adjustment of 尔 must apply to all regions, not just the Chinese regions.~

I know almost if not all Japanese fonts have the two parts connected, but seems like it's a design choice rather than the standard. On the other hand, most of the Chinese fonts show 尔 as disconected, but some are connected like Yuanti SC/TC and Hiragino Sans SC/TC.

Character Glyph Name Notes
U+5C14 尔 uni5C14-CN Consider restoring the CN glyph, and consider restoring or adding new CN glyphs for U+40A7 䂧, U+47E2 䟢, U+4F60 你, U+59B3 妳, U+5F25 弥, U+6CB5 沵, U+72DD 狝, U+73BA 玺, U+73CE 珎 (GE), U+7962 祢, U+79F0 称, U+8999 覙, U+8D82 趂 (GE), U+8FE9 迩, U+9268 鉨 (GE), U+9FDF 鿟, and U+9FED 鿭 (list may not be be exhaustive) whereby the 尔 component does not connect 小 with the horizontal stroke above

~Moving on, in some kind of component relevancy, I think that the components of 欠 may need to be adjusted and unified to the CN form. Some Japanese fonts show the 人 part disconnected from the top part, and that would save some glyphs if Source Han Sans and Serif can follow the CN form for all regions.~

The illustrations below shows the different typefaces for 尔 and 欠, and the ones circled shows the deviancy from the respective "standard" form, highlighted in different colours.

Sans version Screenshot 2022-08-16 at 16 36 58

Hiragino Sans JP/SC/TC comparison (they're almost the same design wise) Screenshot 2022-08-16 at 16 55 05

Serif version Screenshot 2022-08-16 at 16 37 07

Blue is the Japanese fonts, yellow is the Simplified Chinese fonts, green is the Traditional Chinese fonts.

Marcus98T avatar Aug 16 '22 08:08 Marcus98T

I have an update. The Japanese Kanji standards document which I found on the internet, has deemed the stroke-touching difference to be a design difference rather than a regional difference.

Screenshot 2023-03-03 at 16 00 34

Which means, while I am still not sure of the outcome until v3 is released, it's not when former employee Dr Ken Lunde used to inform us in advance of such outcomes with regards to component unification or disunification, there is some possibility that 欠 and/or 尔 could still be unified to the CN form for all regions, but I think this is unlikely.

The most realistic outcome I think would be to separate 尔 as a regional difference and 欠 remains separated when v3 rolls around, unless something is changed.

Marcus98T avatar Mar 07 '23 06:03 Marcus98T

There's just so many issues at this point that the older ones have been buried, but I found a relevant issue about a mainland Chinese user complaining about the 尔 component not complying with the PRC glyph conventions:

  • #255

So I guess with the next update, 尔 will be disunified between JP and CN, and 欠 remains separated, but the merging of non-essential regional differences in other components can make up for this.

Marcus98T avatar Jun 29 '23 01:06 Marcus98T

Here is a table of characters to disunify the 尔 component, if it actually went ahead a long time ago. I think it may have been already done, but I would like to suggest some adjustments and Adobe can verify the table. Also, I would like the shape and proportions of the new CN 尔 to match JP if possible, of course with the two parts being disconnected.

Screenshot 2023-07-21 at 22 31 37

LEGEND Orange - JP is using CN glyph, replace with a proper JP glyph based off Kozuka Gothic Yellow - Adjust CN glyph Light Green - The v1 glyph for CN (or glyph based off Kozuka Gothic for JP) to restore to Cyan - CN glyph is technically compliant, but I suggest to adjust the other parts Blue - Create new CN glyph based off the JP glyph Magenta - Revert glyph to v1 CN glyph Purple - Unsure

Unicode Character Action Taken (JP) Action Taken (CN)
U+4F60 JP is using CN glyph, suggest to include glyph based on Kozuka Gothic and make it default locale for JP and KR, see #425. Adjust glyph
U+59B3 - Adjust both CN and TW glyph.
U+4450 - Unsure, see Unicode charts.
Screenshot 2023-07-21 at 22 43 42
U+47E2 - Adjust glyph
U+9268 - Create new CN glyph based off JP glyph, depends on whether the left 金 radical is unifiable to the JP forms, see #432. This is a GB Extension character.
U+9FED - Adjust glyph
U+73BA JP is using CN glyph, suggest to include glyph based on Kozuka Gothic and make it default locale for JP and KR, see #425. Adjust glyph
U+6CB5 Not in Adobe-Japan1, keep it if there's enough space Create new CN glyph based off JP glyph.
U+7962 - Adjust glyph
U+9FDF - Adjust glyph
U+8FE9 - Adjust glyph
U+5F25 - Adjust both CN and HK glyph.
U+72DD - Adjust glyph
U+60A8 - While the top part is compliant, I suggest to adjust the bottom 心 part to match JP form. Also applies to the TW glyph.
U+3CFD - No action needed, glyph is compliant.
U+37DC - No action needed, glyph is compliant.
U+7315 - No action needed, glyph is compliant.
U+40A7 - Adjust glyph
U+73CE - Create new CN glyph based off JP variant glyph (uni73CEuE0101-JP), v1 CN glyph is not compliant either, see #432. This is a GB Extension character.
U+8999 - Revert to v1 glyph.
U+8D82 - Restore v1 CN glyph. This is a GB Extension character.

And it is likely that HK could follow the new CN forms because TW is also following CN forms too, and I think the HK standard (according to the Pingfang font as of macOS Monterey, not the Unicode charts) is inconsistent in whether 尔 should be connected or not.

Screenshot 2023-07-21 at 23 19 07

Marcus98T avatar Jul 21 '23 14:07 Marcus98T