From: Peter Kirk (email@example.com)
Date: Thu Jan 15 2004 - 05:53:52 EST
On 14/01/2004 16:10, John Delacour wrote:
> What is input by a certain keyboard layout is entirely at the
> discretion of the designer of the layout. If I want to type small
> Greek Alpha with perispomeni, I can use the hex input to type either:
> a) option + 1fb6
> a) option + 03b10342
> The resultant character will be identical.
Correction: it *should* be identical. Unicode specifies that the
rendering of canonically equivalent sequences should ideally be
identical, but not that it must be. And in practice on many systems it
is not. For example, it is not in Windows if you don't have Office 2003
> If I want to type Lao, I get a keyboard layout to do so. Once I have
> it, I can type Lao in any editor that supports Unicode. That's all
> there is to it.
Yes and no. Yes, the editor should be able to process the text, and save
it or transmit it to another system. But if you want to display or print
the text, you need a font which supports the character, and you also
need a rendering system which supports it. To illustrate the problem,
try typing Hebrew characters into your favourite Nisus. Even if you have
a font, you will not get correct RTL rendering - I have just heard from
a Nisus rep an explanation of why.
> If parts of the new MS suite have difficulty rendering any particular
> sort of text, then that would be hardly surprising for applications
> that haven't even been released yet and any rough edges, if they
> exist, would be tidied up in due course.
Rendering of complex and bidi scripts is not a rough edge to be tidied
up in due course. It is a major system enhancement with considerable
complexity. We can hope that MS will do this to MacOffice (and MS has
been under great pressure to do so, including anti-trust action and a
government boycott in Israel), but no indication has been given that
this can be expected soon.
-- Peter Kirk firstname.lastname@example.org (personal) email@example.com (work) http://www.qaya.org/
This archive was generated by hypermail 2.1.5 : Thu Jan 15 2004 - 06:27:11 EST