Updating TW glyph shape to CNS11643 Amendment 1
NightFurySL2001 opened this issue · comments
See adobe-fonts/source-han-sans#454 for details. It seems that Serif has more strictly following CNS11643 glyphs.
Action list for doing CNS11643 Amendment 1:
| Char | Modification | Action | Related info | |
|---|---|---|---|---|
| 1 | 罕 | Fix 儿 to touch | No action | See also item 19 |
| 2 | 承 | Right component cross over | Remove TW, remap TW/HK to JP | Ignorable difference in HK? |
| 3 | 新 | 2nd horizontal stroke shorter than 3rd horizontal stroke | Modify TW glyph | Modify HK too to 2nd longest |
| 4 | 踴 | Hook at 勇 | No action | Using HK |
| 5 | 謄 | 誊 start at 2nd horizontal stroke | No action | Using CN, see #39 (comment) |
| 6 | 璽 | Hook at 爾 | No action | Using JP |
| 7 | 櫬 | See 新 | No action | See item 3 |
| 8 | 襯 | See 新 | Modify TW glyph (1st/2nd horizontal stroke is different)? | See item 3 |
| 9 | 旲 | Change 大 to throw stroke | Remove TW, remap to JP | |
| 10 | 汩 | Flatter 曰 | Modify TW glyph? | |
| 11 | 迗 | Separate dot stroke on 天 | Modify TW glyph/No action? | |
| 12 | 枲 | Separate 口 and 木 | No action | Corrected by accident |
| 13 | 奊 | Change 大 to dot stroke | Add TW glyph | |
| 14 | 埜 | Change 林 to dot stroke | Remap TW to HK | See also item 20 |
| 15 | 掰 | Change 分 to dot stroke | Modify TW glyph, map HK to new TW | See issue adobe-fonts/source-han-sans#340 |
| 16 | 椕 | Change 分 to dot stroke | Remove TW, remap TW to HK | See also item 15 |
| 17 | 萒 | Change bottom 兄 to 允 | Remove TW, remap TW to HK | Double check HK glyph to match CN aesthetics? |
| 18 | 楶 | Change 欠 to throw stroke | Modify TW glyph | |
| 19 | 滘 | Fix 儿 to touch | Modify TW glyph (left 丿 of 儿 is not touching)? | See also item 1 |
| 20 | 榃 | Change 林 to dot stroke | Remap TW to HK | See also item 14 |
| 21 | 葖 | Change 大 to throw stroke | Modify TW glyph, add current TW as HK | Double check with HK? |
| 22 | 獡 | Change bottom right | Remove TW, remap TW to CN | May use current TW for U+2486F |
| 23 | 嬔 | Fix Big5 bug, change 免 to 兔 | No action | Corrected |
| 24 | 瑿 | Change 又 to throw stroke | Modify TW glyph, add current TW as HK | Double check with HK? |
| 25 | 麇 | Change 禾 | Modify TW glyph, add current TW as HK | Double check with HK? |
| 26 | 嬮 | Change structure | Modify TW glyph | |
| 27 | 儭 | See 新 | Modify TW glyph (1st/2nd horizontal stroke is different)? | See item 3 |
| 28 | 檶 | Change structure | Modify TW glyph | Acceptable for HK |
| 29 | 齌 | Change 火 to dot stroke | Modify TW glyph | |
| 30 | 嚫 | See 新 | No action | See item 3 |
| 31 | 瀙 | See 新 | No action | See item 3 |
| 32 | 礡 | Fix Big5 bug | No action | Corrected |
| 33 | 鷍 | Change 木 | Add TW glyph | |
| 34 | 巕 | Fix Big5 bug, change top from 艹 to 䒑 | Modify TW glyph | See issue adobe-fonts/source-han-sans#169 and adobe-fonts/source-han-sans#369, also need to fix 女 |
| 35 | 鱍 | Change 又 to dot stroke | Remove TW, map TW to CN | Using CN |
| 36 | 礴 | Fix Big5 bug | No action | Corrected |
(typo for 嚫)
In my assessment, the list comprises glyphs that fall into two distinct categories, which in my opinion would be best addressed separately:
(1) Changes that involve differences at the stroke or component level. For example, the transformation of ㇏ to 丶 in the character 旲, or the substitution of 口 with 厶 in 䆓.
(2) Changes that reflect variations in design but have been altered to ensure consistency within the CNS reference glyph set. For instance, determining whether the longest horizontal stroke in the 亲 component should be 木, or whether the last (and second to last) stroke should touch the upper component in ⿱㓁 and ⿱穴.
I'd like to add my opinion regarding the second category.
Let's talk about characters composed with ⿱㓁 or ⿱穴 first. I will summarize the current situation in the table provided below:
Please understand that I have no issue with these discrepancies and appreciate the shared design between TW and HK. To me, TW and HK standard glyphs are considered as references that do not necessarily require strict adherence. Thus, when it comes to factors like stroke connections, I’d prioritize aesthetics over the render found in the reference glyph. Hence my points:
- ⿱㓁 and ⿱穴 have different origins (⿱㓁 comes from 网, while ⿱穴 is 穴 and not ⿱㓁[网] with a dot on it). Therefore, their designs do not necessarily have to be identical.
- While I do not oppose unifying the connection strategy of ⿱㓁 and ⿱穴, I fail to comprehend why design A was chosen by the standard body. Personally, I believe form B or C would result in a more aesthetically pleasing design. Moreover, adopting the A form appears even more challenging for heavier weights of the Serif style.
If I were to make a decision, I would likely opt for design B or C, rather than design A. But once again, I am completely fine with these discrepancies. The problem is actually uni7F55-TW 罕 (in Serif) and uni5E18-TW 帘 (in Sans) not in line with the other glyphs.
Therefore, I would like to suggest modifying the uni7F55-TW 罕 to adopt the "both disconnected" form (form C) instead.
Second, for characters composed with the 亲 component, for a more complete picture, I'd like to add that there are actually twelve codepoints or 21 glyphs to look into for the sake of design consistency.
| # | Codepoint | Character | TW | HK |
|---|---|---|---|---|
| 1 | U+512D | 儭 | uni512D-TW | uni512D-HK |
| 2 | U+567A | 噺 | uni567A-HK | |
| 3 | U+56AB | 嚫 | uni56AB-TW | uni56AB-HK |
| 4 | U+65B0 | 新 | uni65B0-TW | uni65B0-HK |
| 5 | U+6AEC | 櫬 | uni6AEC-TW | uni6AEC-HK |
| 6 | U+6FB5 | 澵 | uni6FB5-HK | |
| 7 | U+7019 | 瀙 | uni7019-TW | uni7019-HK |
| 8 | U+85AA | 薪 | uni85AA-TW | uni85AA-HK |
| 9 | U+85FD | 藽 | uni85FD-TW | uni85FD-HK |
| 10 | U+896F | 襯 | uni896F-TW | uni896F-HK |
| 11 | U+89AA | 親 | uni89AA-TW | uni89AA-HK |
| 12 | U+21905 | 𡤅 | u21905-HK |
Note that 罕 specifically in the CNS Amendment 1 change the second to last stroke of 㓁 to lightly touch the top; otherwise I don't see any significant changes between the two glyphs.
This is as reported by an user on the CNS11643 comment page:
It is noted that the officials do mentioned small deviations are acceptable in this case depending on the font style.
帘 in Sans is a must-fix for TW since it doesn't follow other 穴 component.
For 亲 component, there is no definition on which stroke in the component should be the longest in HK. Looking at the reference glyphs, it do seems that there is some changes to make the 3rd horizontal stroke the longest:
[Funnily enough, some HB characters (新, 親 and 薪) seem to have differences between Regular script and Song in the CCLI reference. Maybe it's not modified thoroughly.]
Note that 罕 specifically in the CNS Amendment 1 change the second to last stroke of 㓁 to lightly touch the top; otherwise I don't see any significant changes between the two glyphs.
This is as reported by an user on the CNS11643 comment page:
Oh dear. While I appreciate consistency within a (reference) font, this seems have gone a bit too far to me.
Note: the EDB reference glyph in hk for 親 and 新 both have the second horizontal stroke way longer than the third.
About the 亲 component, personally I am for the change (i.e. adjusting the 3 horizontal strokes of the 亲 component such that the length of 3 >1 >2) because of better consitency, and the style is (mostly) followed by both TW and HK reference glyphs.
By the way, this is how I write it, for what it's worth, which perhaps also explains why I am for the change.
Note: the EDB reference glyph in hk for 親 and 新 both have the second horizontal stroke way longer than the third.
But not 襯, interestingly. While at it, here are the hand-written reference glyphs by the Institute of Language in Education that precedes EDB's computer rendered reference glyphs. The length information was kind of lost during the glyph computerization process, maybe. Or people just don't care as long as they don't look odd or cause confusion.
