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

Possible to restore v1.004 JP glyphs for some characters to replace all locales?

Open Marcus98T opened this issue 2 years ago • 3 comments

Screenshot 2022-02-06 at 04 48 38

Some removed JP glyphs which can be suitable for all regions are generally much better designed, so I suggest to restore the v1.004 JP form and replace the corresponding CN/TW/HK glyphs for all locales, for the following characters (this is a preliminary list for the time being):

  • 俖 (uni4FD6-JP -> uni4FD6-TW)
  • 卟 (uni535F-JP -> uni535F-CN)
  • 叻 (uni53FB-JP -> uni53FB-CN)
  • 叼 (uni53FC-JP -> uni53FC-CN) (although if that can be restored, the last slanted stroke at 刁 (the right component) should be adjusted to make it a little bit more angled, but not too steep angled as in the current CN form)
  • 咔 (uni5494-JP -> uni5494-CN)
  • 咣 (uni54A3-JP -> uni54A3-CN) (although that might need some weight adjustments as the character seems a little thinner compared to the other characters)
  • 嗥 (uni55E5-JP -> uni55E5-CN)
  • 嚉 (uni5689-JP -> uni5689-CN)
  • 奋 (uni594B-JP -> uni594B-CN) (I think the separated 大 component at the top is also acceptable for the CN region)
  • 筘 (uni7B58-JP -> uni7B58-CN) (EDIT 2: Adjust the 口 in uni7B58-JP when restoring)

Marcus98T avatar Feb 05 '22 20:02 Marcus98T

Here is a list for CJK Extension A:

Screenshot 2022-02-10 at 02 59 05
  • 㑞 (uni345E-JP -> uni345E-CN) (I honestly thought CN looks better in balance?)
  • 㗫 (uni35EB-JP -> uni35EB-CN)*
  • 㚈 (uni3688-JP -> uni3688-CN)*
  • 㵮 (uni3D6E-JP -> uni3D6E-CN) (EDIT: There appears to be a stroke mismatch on the bottom right 革 part. See below.)
  • 䊅 (uni4285-JP -> uni4285-CN)*
  • 䊢 (uni42A2-JP -> uni42A2-CN)*
  • ~䍪 (uni436A-JP -> uni436A-CN)~ (EDIT: This JP glyph should not be restored because there's a stroke in the left 羊 component that makes it non-compliant with Chinese standards)
  • 䔊 (uni450A-JP -> uni450A-CN)**
  • 䔤 (uni4524-JP -> uni4524-CN)
  • 䦓 (uni4993-JP -> uni4993-CN)*** (Although honestly CN looks better, just need to lower the 占 a tiny bit to match the v1 JP form)
  • 䦚 (uni499A-JP -> uni499A-CN)***
  • 䦜 (uni499C-JP -> uni499C-CN)*** (The 口 in 吾 should add the left feet and make it a little more centre to 五 if CN glyph is to be kept, otherwise, remove the right feet when restoring the JP glyph)

*This will only go ahead if merging of the 夕 component favours JP/TW/HK forms. If CN forms are picked as the basis for merging, do not restore this character. **This will only go ahead if merging of the 䒑 (as in 豆) component favours JP forms. If CN forms are picked as the basis for merging, do not restore this character. ***If the JP form is to be restored, please remove the right feet from the bottom 口 part as per this issue.

EDIT (2022-02-21): For 㵮 (U+3D6E), there appears to be conflicting information regarding the shape of the bottom right 革 part. Most Chinese fonts (including v2 of Source Han Sans) depict it as one horizontal stroke,

Screenshot 2022-02-21 at 18 15 12

but some other Chinese fonts (including the Unicode reference) depict it as two horizontal strokes.

Screenshot 2022-02-21 at 18 15 17 Screenshot 2022-02-21 at 18 02 01

I suggest if the one horizontal stroke is the most up-to-date, continue to restore the v1 JP glyph AND adjust the 革 part to make it one horizontal stroke. Otherwise, restore the v1 JP glyph as-is.

Source Han Serif has the two horizontal strokes version. Adjust it as well IF one horizontal stroke is the most up-to-date. Screenshot 2022-02-21 at 18 06 45


Below is only for me to keep in view for the time being:

The following list of JP glyphs which on the surface, could have looked okay for all regions, but really shouldn't be restored because that might break Chinese national standards:

  • 㔁 (uni3501-CN)
  • 㤠 (uni3920-CN)
  • 㥦 (uni3966-CN)
  • 㬥 (uni3B25-CN)
  • 㳟 (uni3CDF-CN)
  • 䔵 (uni4535-CN)
  • 䣛 (uni48DB-CN)
  • 䣴 (uni48F4-CN)

And in addition, these glyphs shouldn't be restored unless 戶 and 子 would ever be unified to the JP forms:

  • 㪿

Marcus98T avatar Feb 09 '22 18:02 Marcus98T

I found conflicting glyph information regarding 㵮 (U+3D6E). Someone needs to further check which glyph shape is the most up-to-date. I have edited my second post to reflect this as well. Regardless of what happens, the v1 JP glyph must still be restored. It can be adjusted later as necessary.

Source Han Serif also has this issue as well.

Marcus98T avatar Feb 21 '22 10:02 Marcus98T

I forgot about this, I am currently checking for additional glyphs, but for now, consider restore the v1 JP glyph for 摵 (U+6475, in red) to replace the current glyph (in blue).

摵 comparison

Screenshot 2023-04-02 at 17 14 01

Depends on which glyph looks better.

Marcus98T avatar Apr 02 '23 09:04 Marcus98T