Re: Another --> Re: GtkMovementStep of GtkTextView
- From: Steve Underwood <steveu coppice org>
- To: gtk-i18n-list gnome org
- Subject: Re: Another --> Re: GtkMovementStep of GtkTextView
- Date: Fri, 25 May 2001 09:43:50 +0800
Pablo Saratxaga wrote:
>
> Kaixo!
>
> On Thu, May 24, 2001 at 09:06:48PM +0200, Pablo Saratxaga wrote:
>
> > On Thu, May 24, 2001 at 02:02:55PM -0400, Owen Taylor wrote:
> >
> > > input method level. The main disadvantage is that languages that
> > > otherwise don't require an input method (like Tamil) would then
> >
> > Tamil doesn't require an input method when TSCII encoding is used (it
> > is a visual encoding, similar in concept to Thai encoding);
> > but for ISCII-TML or for unicode an input method is need, the same
>
> Mmh, I realize that in fact there must be two different keyboard layouts for
> tamil; one is the iscii like, where there is no need of input method to
> type in utf-8; the other is the layout used by people typing in tscii;
> there would probably be a need to have an input method to allow typing
> in utf-8 using that layout, that probably a lot of people are used to.
There are more than two keyboard layouts. I saw a word processing
package in India which had about 5 or 6 keyboard layouts for each
language it supported, including Tamil. There was a government sponsored
layout, a "try to be as close as possible to English" layout, a "we
think this is the fastest keyboard for text entry" layout, and some
others. I have no idea how many are in widespread use, though.
If storing all Unicode as normalised Unicode becomes the norm, all Indic
languages would need an input method to do the normalisation before
commital. I think such as step would be a good thing, as normalised
Unicode gives meaningful string comparisions, without the runtime
overhead of normalising before every single comparison.
Regards,
Steve
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]