Adjustment suggestion to the 彗 component for HK


First off, the CN version cannot be shared with JP/TW/HK anyway due to the design difference of ⺕, so it will be ignored throughout the discussion.

## The situation

Standard wise, the last/bottom horizontal-alike stroke in the top-left component 丰 of 彗 is chosen differently between JP, TW and HK.


For JP, a 挑 (Upward horizontal/Tí/Tiāo/Hane) form is used. I'll call it the "JP form".

For TW, a 橫 (Horizontal/Héng/Yoko) form is used. I'll call it the "TW form".

For HK, there are two variants:

- When written in full width (whether there is components above/below it doesn't matter), the stroke will be a horizontal one. Simply put, it is the same as the TW form.

- When there is component(s) left or right to it, the 彗 component is considered "compressed", and the stroke appears slightly slanted upwards. This is a 斜橫, a stroke form found in 毛,七,斗 etc. Its form is still a horizontal stroke, different from a 挑.

## The problem

The problem is that, intentional or not, while Source Han Serif did't adopt the HK-specific 斜橫 form, the stroke of 丰 in 彗 is not consistent. Here is a list of the codepoints concerned:

| # | Codepoint | Character | Current HK glyph | Current stroke form
| --- | --------- | --------- | ---------- | ----------------
| 1 | U+3EF0 | 㻰 | uni3EF0-HK | 橫 (TW)
| 2 | U+439A | 䎚 | uni439A-HK | 挑 (JP)
| 3 | U+5612 | 嘒 | uni5612-JP | 挑 (JP)
| 4 | U+66B3 | 暳 | uni66B3-JP | 挑 (JP)
| 5 | U+69E5 | 槥 | uni69E5-JP | 挑 (JP)
| 6 | U+6AD8 | 櫘 | uni6AD8-HK | 挑 (JP)
| 7 | U+8B7F | 譿 | uni8B7F-TW | 橫 (TW)
| 8 | U+8F4A | 轊 | uni8F4A-JP | 挑 (JP)
| 9 | U+93CF | 鏏 | uni93CF-TW | 橫 (TW)
| 10 | U+218FA | 𡣺 | u218FA-HK | 挑 (JP)
| 11 | U+24AB3 | 𤪳 | u24AB3-HK | 挑 (JP)
| 12 | U+26484 | 𦒄 | u26484-HK | 橫 (TW)
| 13 | U+28BDA | 𨯚 | u28BDA-HK | 挑 (JP)

Show image

I consider this an overlook, as the design of the new HK glyphs are also not consistent (interestingly ,it *is* consistent that the new HK glyphs use 橫 for 彗 and 挑 for 慧). I do not insist that the HK-specific form be followed given the Pan-CJK nature of this typeface and the technical limit of CID fonts, but I would like a consistent form be adopted for the "compressed" 彗 for HK. Thus the following possible solutions:

## Option 1: Use the JP (upward) form

Pros:
- Most economical, least efforts needed. Only 2 redesign and 2 new glyphs.
- The font adopts a similar design choice for ⿰㓞 - the HK reference glyph uses a 斜橫 but Source San Serif uses a 挑 to maximize glyph sharing with TW. Examples:


Cons:
- Glyph forms not sharable with TW.

### A. No action needed (9)

| # | Codepoint | Character | Current HK glyph | Current stroke form
| --- | --------- | --------- | ---------- | ----------------
| 2 | U+439A | 䎚 | uni439A-HK | 挑 (JP)
| 3 | U+5612 | 嘒 | uni5612-JP | 挑 (JP)
| 4 | U+66B3 | 暳 | uni66B3-JP | 挑 (JP)
| 5 | U+69E5 | 槥 | uni69E5-JP | 挑 (JP)
| 6 | U+6AD8 | 櫘 | uni6AD8-HK | 挑 (JP)
| 8 | U+8F4A | 轊 | uni8F4A-JP | 挑 (JP)
| 10 | U+218FA | 𡣺 | u218FA-HK | 挑 (JP)
| 11 | U+24AB3 | 𤪳 | u24AB3-HK | 挑 (JP)
| 13 | U+28BDA | 𨯚 | u28BDA-HK | 挑 (JP)

### B. Redesign needed (2)

| # | Codepoint | Character | Current HK glyph | Current stroke form
| --- | --------- | --------- | ---------- | ----------------
| 1 | U+3EF0 | 㻰 | uni3EF0-HK | 橫 (TW)
| 12 | U+26484 | 𦒄 | u26484-HK | 橫 (TW)

### C. New glyph needed (2)

| # | Codepoint | Character | Current HK glyph | Current stroke form
| --- | --------- | --------- | ---------- | ----------------
| 7 | U+8B7F | 譿 | uni8B7F-TW | 橫 (TW)
| 9 | U+93CF | 鏏 | uni93CF-TW | 橫 (TW)

## Option 2: Use the TW (fully horizontal) form

Pros:
- This just means the HK-V1 form is used for all cases.
- The redesigned/new glyphs are sharable with TW.
- The basic stroke form is still preserved.
- Same new glyph count as option 1.

Cons:
- More modification efforts needed.

### A. No action needed (4)

| # | Codepoint | Character | Current HK glyph | Stroke form
| --- | --------- | --------- | ---------- | ----------------
| 1 | U+3EF0 | 㻰 | uni3EF0-HK | 橫 (TW)
| 7 | U+8B7F | 譿 | uni8B7F-TW | 橫 (TW)
| 9 | U+93CF | 鏏 | uni93CF-TW | 橫 (TW)
| 12 | U+26484 | 𦒄 | u26484-HK | 橫 (TW)

### B. Redesign needed (5)

| # | Codepoint | Character | Current HK glyph | Current stroke form
| --- | --------- | --------- | ---------- | ----------------
| 2 | U+439A | 䎚 | uni439A-HK | 挑 (JP)
| 6 | U+6AD8 | 櫘 | uni6AD8-HK | 挑 (JP)
| 10 | U+218FA | 𡣺 | u218FA-HK | 挑 (JP)
| 11 | U+24AB3 | 𤪳 | u24AB3-HK | 挑 (JP)
| 13 | U+28BDA | 𨯚 | u28BDA-HK | 挑 (JP)

### C. Remapping needed (2)

| # | Codepoint | Character | Current HK glyph | Current stroke form | Remap target
| --- | --------- | --------- | ---------- | ---------------- | ------------
| 3 | U+5612 | 嘒 | uni5612-JP | 挑 (JP) | uni5612-TW
| 5 | U+69E5 | 槥 | uni69E5-JP | 挑 (JP) | uni69E5-TW

### D. New glyph needed (2)

| # | Codepoint | Character | Current HK glyph | Current stroke form
| --- | --------- | --------- | ---------- | ----------------
| 4 | U+66B3 | 暳 | uni66B3-JP | 挑 (JP)
| 8 | U+8F4A | 轊 | uni8F4A-JP | 挑 (JP)

## Option 3: Use the HK form

Pros:
- The HK-specific design is fully acknowledged.

Cons:
- The least economical choice; largest number of new glyphs needed.

### A. Redesign needed (7)

| # | Codepoint | Character | Current HK glyph | Current stroke form
| --- | --------- | --------- | ---------- | ----------------
| 1 | U+3EF0 | 㻰 | uni3EF0-HK | 橫 (TW)
| 2 | U+439A | 䎚 | uni439A-HK | 挑 (JP)
| 6 | U+6AD8 | 櫘 | uni6AD8-HK | 挑 (JP)
| 10 | U+218FA | 𡣺 | u218FA-HK | 挑 (JP)
| 11 | U+24AB3 | 𤪳 | u24AB3-HK | 挑 (JP)
| 12 | U+26484 | 𦒄 | u26484-HK | 橫 (TW)
| 13 | U+28BDA | 𨯚 | u28BDA-HK | 挑 (JP)

### B. New glyph needed (6)

| # | Codepoint | Character | Current HK glyph | Current stroke form
| --- | --------- | --------- | ---------- | ----------------
| 3 | U+5612 | 嘒 | uni5612-JP | 挑 (JP)
| 4 | U+66B3 | 暳 | uni66B3-JP | 挑 (JP)
| 5 | U+69E5 | 槥 | uni69E5-JP | 挑 (JP)
| 7 | U+8B7F | 譿 | uni8B7F-TW | 橫 (TW)
| 8 | U+8F4A | 轊 | uni8F4A-JP | 挑 (JP)
| 9 | U+93CF | 鏏 | uni93CF-TW | 橫 (TW)

That's it. Thanks!