Re: Merging CoreBase into Base

[David Chisnall]

On 12 Aug 2013, at 16:26, Stefan Bidi <address@hidden> wrote:

> (2) I'm working towards making corebase use Unicode (ie UTF-16) internally 
> wherever possible. I believe this is a saner choice than trying to deal with 
> UTF-8.

I find this an odd observation.  UTF-16 is multibyte, so comes with all of the 
same pain as UTF-8, but has the disadvantage that it's almost always larger 
than UTF-16 (most two-byte characters in UTF-16 are also two-byte characters in 
UTF-16).  You also start hitting endian issues with UTF-16, whereas UTF-8 is 
endian-independent.  Finally, UTF-8 is the format that you typically want for 
input or output, as it's well supported by most libc functions, terminals, and 
so on.

David