4.1106 Fixed Width Characters (1/22)

Elaine Brennan & Allen Renear (EDITORS@BROWNVM.BITNET)
Thu, 28 Feb 91 22:54:21 EST

Humanist Discussion Group, Vol. 4, No. 1106. Thursday, 28 Feb 1991.

Date: Thu, 28 Feb 91 11:00:26 CST
From: Michael Sperberg-McQueen <U35395@UICVM>
Subject: fixed width characters

(Apologies to those bored by character-set issues. Stop reading *now*.

I fear there has been a misunderstanding. 1 There seems to me to be no
'knock' in saying that a code which provides dead keys makes possible a
data stream with varying-width characters. I didn't say it was good or
bad. 2 As I understand Bill Tuthill's posting, and the other postings,
and the Unicode document, Unicode does provide dead keys. Since it also
provides composed characters, Mr. Tuthill points out, no one need use
the varying-width encoding unless they wish to. True. This does not
seem to me to affect the issue raised by Pierre Mackay, however, which
is whether Unicode and ISO10646 provide guaranteed fixed-width-character
data streams.

If there are dead keys, they can stack, and character width can vary.
If Unicode provides dead keys, Unicode character width can vary. As Mr.
Tuthill says, some like it that way; some do not. His comments on
implementation difficulties are to the point, though opaque. It would
be nice, however, to engage in technical discussion without invidious
political comments.

Michael Sperberg-McQueen