I’ve been setting up localisation for an application using a combination of tools: the Boost Locale library, and the widely used gettext tools. I’m wanting to work entirely in UTF-8 because it should be suitable for pretty much any language we’re likely to need, and it should hopefully avoid problems of mixing encodings. However, I found that the tools kept falling-back to outdated character sets instead.

Fortunately, there’s a fairly easy solution which I’ll outline quickly in this post. Read more Generate UTF-8 dictionaries using gettext ›