If possible, replace some CN (or TW/HK) glyphs with unreleased JP glyphs #425
Comments
This was referenced Jun 15, 2023
|
Here is a list for the characters within Adobe-Japan1 that are using CN glyphs (or rarely TW/HK glyphs), and they never had a v1 JP glyph (although for Serif, some of them had v1 JP glyphs). They may not be entirely accurate because some glyphs may have been renamed, but I have filtered the characters outside of the Basic Multilingual Plane to make sure there are no mappings to any existing JP glyph. There are 1509 characters within Adobe-Japan1 that only have CN glyphs for Sans (as of v2.004) and 432 characters for Serif (as of v2.002). It is likely the designer, Ryoko Nishizuka, had determined that the CN glyphs look better across different locales. However, I would partially disagree, because there are some minor stroke design differences between JP and CN which can break the consistency of the overall look and feel (especially when using the Japanese/Korean locales), along with what I deem ugly proportions of some CN glyphs. But ultimately, there may be no action taken by Adobe for this issue, so it is for reference only. UPDATE 2023-08-18: Removed 岸 (U+5CB8) from the Serif table following the new release of Serif v2.002. UPDATE 2023-09-22: Added status on whether some characters in Serif had v1 JP glyphs, but the recommendation to restore is already on this page. UPDATE 2024-01-06: Based on my recent email correspondence with someone formerly involved in this project, it was actually the designer who chose the CN glyphs over the JP glyphs. Anyway, I'm sure Adobe still has the original JP glyphs, so this issue can be partially solved in a future major update. Here is the part of their reply that confirms how the process was done when making Source Han Sans: Sans
Serif
|
This was referenced Jul 21, 2023
This was referenced Oct 30, 2023
This was referenced Nov 15, 2023
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
I request to replace some (but not all) CN/TW/HK glyphs, which are the only glyphs for all regions, with unreleased JP glyphs based on Kozuka Gothic because the latter are generally better designed. It also probably will solve some slight discrepancies with regards to the radical component designs.
The reason why I am opening this issue is because even in v1, for some characters, there were no JP glyphs, therefore CN glyphs are the only glyphs used for JP/KR. But a good amount of them are not really well-designed, in fact Changzhou Sinotype did the CN/TW glyphs very quickly, which results in a generally poorer quality of glyphs. Which is why I am asking Adobe to use their unreleased JP glyphs based on Kozuka Gothic and replace the CN/TW/HK glyphs, if possible.
I had gone through the AI0-SourceHanSans and AI0-SourceHanSerif lists and roughly managed to filter out some glyphs that never had JP glyphs. I would very much prefer if about 90% of the Adobe-Japan1 character set for SHS are using JP glyphs, except for 的 (U+7684) and some other CN/TW/HK glyphs which are not too much different from the unreleased JP glyphs.
One example is 狙 (U+72D9), it is a CN glyph, because of the unnecessary CJK radical design changes done by Changzhou Sinotype, therefore it looks a bit ugly.
I also pointed out the 心 component in some glyphs, 書 (U+66F8), 狅 (U+72C5), 甭 (U+752D, on Serif) and 悟 (U+609F), see below. So maybe next time if I find such issues they go to this page.
And yes, it also concerns restoring v1 JP glyphs to replace CN glyphs since there are virtually no parts that require a regional design difference, although I think at this point the list over there may need updating.
For Serif, similar issues apply.
UPDATE: Whoops, looks like I already mentioned this issue earlier in #410, but I will move the content here.
UPDATE 2: I have new insights based on my recent email correspondence with someone formerly involved in this project. This is part of their reply:
It also confirms that Sinotype is no longer involved in the project.
The text was updated successfully, but these errors were encountered: