[gcompris-devel] Re: charset generation

[Sergey V. Udaltsov]

Malcolm Tredinnick пишет:

>On Wed, 2004-02-11 at 11:02, Bruno Coudoin wrote:
>  
>
>>Le mer 11/02/2004 à 00:36, Jody Goldberg a écrit :
>>    
>>
>>>On Wed, Feb 11, 2004 at 12:12:32AM +0100, Bruno Coudoin wrote:
>>>      
>>>
>>>>In gcompris, we have an activity where the kid just hit the falling
>>>>letters. 
>>>>As of today, only ASCII is supported.
>>>>
>>>>We would like to display a char from the charset of the current locale.
>>>>Is it possible to get it or we need to hardcode a list of unicode char
>>>>for each locale ?
>>>>        
>>>>
>>>It sounds like what you really need is a way to find all the
>>>characters a child could type with their current keyboard layout.
>>>The locale is only a proxy to that.
>>>
>>>      
>>>
>>You are correct. We are especially focussed on kids and we may want to
>>avoid complex charaters typings. More, we may need to start with letters
>>easy to type (one key) and then introduce the ones that needs 2 keys
>>(like alt+x).
>>    
>>
>
>But in this case you need to build up the list of possible characters
>yourself, surely? It is not locale dependent, rather it is input-method
>dependent.
>  
>
I seen similar type-training progs for DOS/Windows for English and 
Russian layouts. And this type of training seems reasonable.
The first problem you'll encounter is, typical, the issue with 
non-xkb-enabled servers. With proper xkb I am pretty sure the task is 
relatively easy to solve - but what if you dont have xkb?

Sergey