Re: Yet Another I-Dotting Proposal

From: John Cowan (
Date: Mon Aug 11 1997 - 17:22:38 EDT

Jörg Knappen scripsit:

> Now you have a style sheet, demanding that the last names of authors are to
> be set in caps and small caps. No rich text format can help you guess the
> correct form of LOWERCASE I vs. LOWERCASE I WITH DOT.

Actually, that case is the *less* critical one, because you can
correctly encode Turkish "i" as LATIN SMALL LETTER I plus
COMBINING DOT ABOVE. As the Unicode Standard says at p. 6-7,
this sequence looks just like a SMALL I alone.
A distinct SMALL I WITH DOT character is needed only to
make 1-1 case mapping possible, as it is the lowercase form

The really critical need is for a capital form of DOTLESS I.
This would be, as I have said, glyphically identical to
CAPITAL I and would represent yet another place,
like the three Ds with stroke (U+00D0 ETH, U+0110 D WITH STROKE,
U+0189 AFRICAN D), where different lowercase forms
have a glyphically identical capital form but are given separate
capital-letter character codes. Like these cases, CAPITAL
DOTLESS I permits the correct display without losing case reversibility.

John Cowan
			e'osai ko sarji la lojban

This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:36 EDT