Character Sets

Petr Bříza introduces character sets for websites and web applications
in his article Character Sets in Practice.
Alongside the basics of character encodings and character sets, he debunks the myth that
Windows-1250 is a “nonstandard” encoding.

I'd already lean toward using Unicode in a form such as UTF-8,
which supports a wide range of characters and is also widely used.
For a multilingual site or application,
I wouldn't hesitate to use UTF-8.

One thing I missed in the article is that you can change the encoding in the web server's output,
for example with a front-end filter. That approach, together with setting the appropriate header
to announce the document's new encoding, is less common but has been used
successfully.