From: Peter Constable (petercon@microsoft.com)
Date: Mon Apr 26 2004 - 10:47:17 EDT
> >I like to make a proposal for 2 Romanian characters to be added to
the
> >Unicode / ISO 10646-1 (Basic Multilingual Plane) character set.
> >
> Similar problem in Lithuanian: we need 35 (!) accented letters. Our
> official proposal was not accepted.
Because the encoding *principles* that Cristian is looking for are
these:
- text elements with combining marks are represented as
dynamically-composed sequences (TUS4 p. 20)
- exceptions to the above (and other design principles) are permitted
only to ensure round-trip convertibility with pre-existing encoding
standards in wide usage as of 1993 (TUS4 p. 22)
The challenges created by having both precomposed and decomposed Latin
representations are something many of us have to contend with. Adding
new precomposed forms will not make those issues go away, and it will
create other problems, some of them quite serious insofar as it affects
normalization. Because of Unicode's stability policy wrt normalization,
and the above principles, new precomposed forms for Romanian or any
other language will not be added. That's just the way it is.
Peter
Peter Constable
Globalization Infrastructure and Font Technologies
Microsoft Windows Division
This archive was generated by hypermail 2.1.5 : Mon Apr 26 2004 - 11:19:29 EDT