Re: [gtk-i18n-list] Unicode PUA supporting issue in gtk+/pango

[mpsuzuki hiroshima-u ac jp]

On Sat, 31 Dec 2005 11:37:23 +0800
Zhe Su <james su gmail com> wrote:

>  I'm not a familiar with this area. So please excuse me, if I'm
>wrong. I was heard that they had proposed the pre-composed Tibetan
>glyphs to Unicode long time before, but got rejected. I got such
>information from an officer of PRC government.

Thanks. I searched unicode-ml and found PRC proposed
precomposed Tibetan charset: "BrdaRten" 956 characters
to BMP (U+A500-A8BB) on 2002. From the size of charset,
I suppose BrdaRten is "set A".

http://www.unicode.org/mail-arch/unicode-ml/y2002-m12/0101.html
http://std.dkuug.dk/jtc1/sc2/WG2/docs/n2558.pdf

Yet I could not find the conclusion notes why it was
rejected, but from the discussion (sorry for google
cache), the uncomposed Tibetan charset was already
included (and it can cover whole of BrdaRten charset
'theoretically' - in the rejecters' viewpoint), it
might be the reason why new Tibetan charset was
rejected.

http://72.14.203.104/search?q=cache:XbJkML8I8rwJ:std.dkuug.dk/JTC1/sc2/WG2/docs/n2553.pdf+BrdaRten+%22set+B%22&hl=ja
http://66.102.7.104/search?q=cache:JpF_pjpcOJYJ:anubis.dkuug.dk/jtc1/sc2/wg2/docs/n2668.doc+BrdaRten&hl=ja

I was wrong in the point that PRC had proposed
the precomposed Tibetan charset, but I cannot
expect as they will be included in SMP etc.
So, still the classification of the charset is:

> B. The character won't be included in Unicode,
>    because it violates Unicode character inclusion
>    policy.

# BTW, don't take me as if I agree with the policy.
# I'm just saying as "they will reject in future,
# as they did against non latin characters".

>  There is a group in Chinese Academy of Sciences working on the
>Tibetan support in Linux. One of their member told me that the quality
>of available Tibetan compose engines can't meet their requirement.
>They didn't just say pango. So they decided to use PUA to store the
>pre-composed Tibetan glyphs. Then all applications should support such
>approach, because their input method can only input Tibetan chars by
>using the code points of those pre-composed glyphs.

Thank you for notice. Their precomposed Tibetan
charset is developed by theirselves? They adopted
PRC's precomposed Tibetan charset A or B? If they
are published, I want to obtain a copy. I think,
there might be users who want to use precomposed
Tibetan charset without adoptation of whole PUA
codepoints definition in PRC.

Regards,
mpsuzuki