> John Cowan began this discussion by asking how U+02BB was chosen, but
> I don't know that he intended to convey that U+2018 would have been
> better (though I don't really know for sure).
I had no such intention. When I asked "how was it chosen?", I meant
to inquire how it was chosen, and nothing more.
FWIW, Cyberbit doesn't implement U+2018 either.
> Please note that, as Ken pointed out, U+2018 is intended as
> punctuation. If you use that as a word-forming character, a browser or
> other app might give you undesireble behaviour (for, say, word/line
> breaks, hyphenation) becuase it is making assumptions about the
> character semantics.
In fact, I think that is unlikely. U+2018 is the preferred character
in English words like "isn't" and "won't", and it is wildly unlikely
that any browser or other text-manipulator would line-break or
hyphenate near it.
(Of course, most English apostrophes are currently U+0027, but that is
a heavily overloaded character and is semi-deprecated.)
> Since U+02BB is intended for "letters of
> alphabets", it is probably a better choice in the long term, even if
> there is narrower implementation for it in fonts at the present time.
Yes. For example, Java programs written using Hawai'ian identifiers
would not work with U+2018, since it is not a letter, but would work
with U+20BB.
-- John Cowan cowan@ccil.org e'osai ko sarji la lojban.
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:40 EDT