From: Richard Wordingham (email@example.com)
Date: Sun Jun 11 2006 - 13:16:00 CDT
Philippe Verdy wrote on Saturday, June 10, 2006 at 2:55 AM
Subject: Re: UTF-8 can be used for more than it is given credit
> Now I am wondering if there is any really valid reason why a combining
> ypogegrameni should turn to a plain iota letter. I now think that such
> case is invalid for all locales, except when applying a change of
> orthography between Ancient Greek and Modern Greek.
Have you read the account at
> So may be the best solution ...
... might be to define the appropriate mappings for the locale 'el' (Modern
Greek). Or perhaps the current mappings should be restricted to the locale
'grc' (Ancient Greek). (Someone's mooted the idea of moving locale specific
case mappings from TUS to the CLDR.) I'm not sure how one calls up
multiple, simultaneous locales. What's appropriate in Sydney or North
London might not be appropriate in Cambridge. It's not a problem if all
text is correctly language-tagged and the application senses the tag when
(and only when) appropriate, but I would not be surprised if most tagged
text were incorrectly tagged.
> is to completely remove the rule that turns ypogegrammeni to a simple iota
> within the simple or full casemappings
To do that completely would necessitate a capital adscript iota. Objections
I can see to adding:
xxxx;GREEK CAPITAL PROSGEGRAMMENON;Mc;0;L;;;;;N;;;;0345;
1) it might increase rather than lessen confusion
2) it might lead to demands for
yyyy;GREEK SMALL PROSGEGRAMMENI;Mc;0;L;;;;;N;;;;0345;
to distinguish the small adscript iota glyph from the subscript iota glyph.
It would be a reincarnation of what is now
3) it would complicate the rules for titlecasing - it would no longer affect
just the first character after a word break!
Is there anywhere we can read the arguments rather than repeating them ad
This archive was generated by hypermail 2.1.5 : Sun Jun 11 2006 - 13:35:24 CDT