Change default forms to follow Unicode forms instead of JIS forms?

[aaronbell]

Starting this issue to create a public location to discuss.

Morisawa engineering team thinks this is about the difference between the JIS standard and Unicode standard. Besides the Yen mark that the customer mentioned, there are other characters following the JIS standard which should be changed if we tend to shift to Unicode standard.

Top line is the JIS version, bottom is Unicode.

@davelab6 response:

I'd like to fix the yen misencoded as backslash as soon as possible because otherwise when we fix it we will break documents using it.

I don't recommend changing the kanji or punctuation designs, only the Unicode values of needed (like yen sign). The image was a bit blurry and I couldn't see that information clearly.

The aim for Google is document portability; that means keeping the glyphs the same but using correct Unicodes where they are very obviously wrong, only yen looks like that to me in this table.

[davelab6]

The last 2 cols of the 1st & 2nd rows might be worth changing.

[aaronbell]

Looking at the mockup, the two that are explicitly problematic design-wise are 005C (yen) and 203E (overline). The remaining punctuation appear to just be positional / sizing. I'm a bit confused by the unicode form shown for 2225 as the codepage shows something that looks like the JIS version.

For the Kanji, I don't think these design variants are necessarily "wrong" in the eyes of Unicode as the variants are listed for these codepoints. For example:

My general sense is that as long as they're not unexpected or incomprehensible to the user (and I think in this case, the JIS version would not be unexpected), they're fine to leave them as is.

[davelab6]

Does these bugs occur in BIZ UD Mincho too?

By overline, do you mean the 2nd ~ on the top row?

Since yen and that are encoding bugs, and GF is committed to Unicode encoding, I'd like to ask you/TN to prioritize a v1.003 release that fixes these ASAP, so that the fix doesn't disrupt too many Workspace user documents.

[aaronbell]

Yes these occur in UD Mincho as well.

overline
Yup.

I’ll follow up and find out the status of these.

[davelab6]

Thanks!

[hiroyuki-komatsu]

I'm wondering the code point of the following character. Is it U+6634 (昴), U+663B (昻), or others? If you have a high resolution image for those code points, would you please share it here?

For U+5307 (匇), J3-2E6C is the sample glyph for Japanese, and others are for Simplified/Traditional Chinese. So if we should pick one of them, I'd prefer Japanese glyphs for UD fonts.

• https://unicode.org/charts/PDF/U4E00.pdf

For U+FA24 (﨤), The Japanese glyph has two dots.

Practically speaking, most people may not care about the differences so much. On the other hand, if we can choose it I'd suggest Japanese glyphs.

Thanks,