Hello,

In reference to Matz's comment below:

>>Yukihiro Matsumoto wrote:
>>    
>>
>>
>  
>
>>>>> > |I want to extend the question - are these methods now implemented with respect
>>>      
>>>
>>>>> > |to "Unicode Standard Annex #21 - Case Mappings" using the mapping information in
>>>      
>>>
>>>>> > |"CaseFolding.txt" (UCD)?
>>>      
>>>
>>>>> > Not yet.  Maybe in the future.
>>>      
>>>
>>>> I think it will be a good idea to have these functionalities as add on
>>>> libraries. All these Unicode related attributes need a lot of data (it's not
>>>> only case mapping), and not everybody needs it.
>>    
>>
I just wanted to say:
Please do consider making unicode dynamically loadable.

Yesterday, I started looking how the 1.2Mb ruby19 interpreter binary could be put on a diet.
For reference, the 1.6.8 ruby binary is about 600Kb.  Turns out that over 100Kb
of the new binary is composed of "Unicode related attributes" defined in the core
source file "unicode.c"

- brent