I agree with Jungshik that U+76F4 (straight) is possibly the case where unification went farthest in the sense that it's the case where average modern readers in various areas might be most (1) confused if they see the glyph variant they are not used to.
(1) 'most confused' should not be misunderstood to be very confused; for most other cases cited often, e.g. the grass radical, the bone radical/character,..., it's difficult for the users to recognize the difference in running text. If you want to give an example of where 'unification fails', then you have to really be precise and speak about the application case. In general, unification just works, that's how it has been designed. Regards, Marti.n At 16:59 01/12/21 -0500, Jungshik Shin wrote: >On Fri, 21 Dec 2001, Suzanne M. Topping wrote: > > > The two examples I dug out of various ongoing email debates etc. are > > below: > > > > The traditional Chinese glyph for "grass" uses four > > strokes for the "grass" radical, whereas the simplified Chinese, > > Japanese, > > and Korean glyphs use three. But there is only one Unicode point > > for the > > grass character (U+8349) regardless of writing system. > > I can't say about Japanese or simp. Chinese, but in Korea >the number of strokes for the 'grass' radical (U+8279) is sometimes three >and other times four. Every Korean with the minimum knowledge of Chinese >characters knows that it can have either four strokes or three strokes. >It's just a matter of taste of font designers, individuals, etc. However, >four stroke version is more common and I think is considered 'canonical'. >Personally, I always (well, like most other Koreans, I rarely use Chinese >characters with a pen/pencil these days) use four strokes. >(ref. http://211.46.71.249/handic/index.htm) > > I've just found that there are three grass radicals >encoded in CJK Radicals supplement block: U+2EBE, U+2EBF, U+2EBF >in addition to the _full_ form at U+2F8B. > > > > Another example is the ideograph for "one," which is different > > in > > Chinese, Japanese, and Korean. > > Did you mean U+58F9 ? > > > > I have been told that neither of these are valid examples, for various > > reasons. > > I guess so. > > > > I very much want to include a legitimate example of a character which > > displays using different glyphs in various character sets, and am hoping > > that one of you brilliant people out there can send me one ASAP, so I > > can finish this blasted paper and go home to grab a glass of eggnog. > > I'm sorry I can go to a great length to show why CJK unification is not >so much a problem as some people have tried to make it look >(e.g. <http://www.xfree86.org/pipermail/i18n/2001-October/002526.html>, > or <http://www.xfree86.org/pipermail/i18n/2001-November/002639.html> >), it's rather hard to give an example you need. > > One potential candidate is U+76F4 (straight). I can't say that the glyph >given for U+76F4 in TUS 3.0 code chart (on p.677) is familiar >to me (although I have little problem recognizing it) . A few years >ago a Japanese sent me an ASCII rendering (as in TUS 3.0 or 2.0) and >challenged me whether I could recognize it. He cited it as an example of >inappropriate unification of traditional Chinese character and Japanese >Kanji. However, this may not be an valid example, either. If you look it >up at http://140.111.1.40 (Chinese character variant dictionary compiled >under the auspice of MoE of Taiwan), the canonical glyph listed for the >character is the same as the glyph on p. 886 of TUS 3.0 (the radical of >the character meaning 'straight' is the character meaning 'eye' U+76EE >with three additional strokes) which is certainly more familiar to me >and, I believe, to Japanese. Therefore, his assertion that traditional >Chinese glyph for U+76F4 is different from Japanese counterpart does not >hold up. As Thomas Chan once remarked on this list, Japanese objection to >CJK unification would have been much smaller if more canonical >glyphs (as listed in the aforementioned variant dictionary) have been >used in TUS 3.0 table. > > Jungshik Shin > >