Re: [ros-general] Ideas (by Paulo Correa Silva)

[robert K.]

KJK::Hyperion wrote:
>   At 01.47 22/02/2004, you wrote:
>  
> >   • Indian keymaps: 
>  
>  [...]
>  
>  from what I've heard, Indian support is going to be a bitch. Unicode 
>  support for Indian is badly designed - or so I'm told - and not able to 
>  represent all Indian text. Operating systems where file names are binary 
>  strings (like most POSIX systems) aren't going to be affected by this. 
>  We, being Windows, will receive the equivalent of a full frontal impact 
>  with a semi truck. Either we'll make liberal use of the user-defined 
>  Unicode range (hoping not to tread over other uses of this range - 
>  Microsoft Interix, IIRC, uses characters in this range to represent 
>  characters that would be otherwise illegal in file names on Windows, 
>  like ":") or we'll be left behind 
Explain me more about indian. Which charset is affected?  Hindi, 
Devangali, unknown?
I know there're some scripts in the world which are context sensitive. 
AFAIK they have one code point for each letter variation. So this should 
be 'just' a problem of correct input mehtod editors....

What you tell us is that several code points are missing :-?
AFAIK UCS-2 / BMP (basic multilingual plane) is rather compleete. There 
are however very rare or died out scripts missing like Keilscript.

So what's exactlY?