Somebody pointed me to your article[1] about the characters that are allowed in class names in CSS. You correctly deduced that é (e with acute accent) is allowed, but your article may lead people to believe that letters in other scripts, such as Chinese or Arabic, are *not* allowed.

I readily admit that the specification is difficult to read, but it does say that all characters outside the ASCII range are allowed. In section 4.1.3 there is the phrase “ISO 10646 characters U+00A1 and higher” and in section G.2 is says that the grammar only shows code points up to (octal) 0377, but that in fact everything up to (octal) 04177777 is allowed.

[1] http://happygiraffe.net/blog/articles/2007/12/19/cliches-are-hard

@charset "<IANA defined charset name>";
@charset "UTF-8";

it belongs at the very beginning of the document. (more)

For the HTML a charset can be added as well into meta headers. This helps to read documents when they were saved and are not provided by the server any longer:

<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />

does that job. But all this is only a very slight addition to your great article.

@codingvista: It depends on the server side language. Certainly Java and Perl allow you to use Unicode in identifiers, possibly Ruby too. I think PHP is the only severely Unicode crippled language at this point.

But in general, this post isn’t really about those specifics. It’s more about: “How do I get a proper answer to my web development problem.” And for that, you do need to go to the specs.

I haven’t time to read the whole lot so you may have already mentioned this but if you reflect CSS id’s/class names in server side code, and there are good reasons why you’d do this, then using complex characters would be a nono.

Just my .5p

w://