Boost logo

Boost :

Subject: Re: [boost] [Boost.Locale] New Release
From: Artyom (artyomtnk_at_[hidden])
Date: 2010-03-13 15:09:08


<Snip>
> I really don't see the point in reinventing a wheel here.
> The code is there, it's well tested and maintained, feature
> rich and widely deployed. What's missing is a better C++
> interface, and that's exactly what I expect from
> Boost.Locale.
<Snip>

Just to provide more points for Andrey's note.

ICU today is de-facto standard Unicode/localization library. Even if you take
a look on CLDR you will find references to ICU as **the** library that
uses CLDR.

The biggest ICU issue is that it was designed after Java interface.
icu4j feels quite natural in Java world but icu4c API is total disaster,
it is like Java translated line by line to C With Classes... (Even
hardly say C++ because it does not use at all: STL, Exceptions, RTTI)

Also, please, read this article:

   Unicode in 2009? Why is it so hard?
   http://art-blog.no-ip.info/cppcms/blog/post/43

I compare major Unicode toolkits available in may languages...
Unfortunately every each one of them fail on quite simple task.
So I think creating full featured Unicode/Localization library
for Boost would be very hard when so many fail in same place.

Artyom

      


Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk