There are issues if the stubs you use in your URLs only allow a-z and 0-9. You could just as easily escape each character in the stub and preserve the title of the post/question you're representing. This is a problem with Stack Overflow though, not the name of the language (C#).

Wikipedia has this issue solved. You can use any Unicode character in their titles and they get used as the stub by escaping those characters. Anyway, it's not like the whole world only uses the English alphabet. This should be handled properly.

C Sharp always seems to me that it should be spelled C♯, not C#. But then I’m coming from a music background.

Yes. The technical term for # is "octothorpe", so if we follow the "comega" approach for C#, that gives us "coctothorpe", a very ...unpretty term.

Far better to type C♯. Or, if you're the one naming a new language, to look for a better name.

It’s hard enough to persuade people to type curly quotes and proper accents even when it’s easy to do (e.g. on OS X). Chances of them learning to type a ♯? Minimal.

On box art and such it is rendered that way: http://en.wikipedia.org/wiki/C_Sharp_(programming_language)#...

I've always had trouble not reading C# as C-pound.

Or as a friend of mine use to say when she started out learning .NET "c hash".

PS. She is from the UK.

Probably has more to do with being from Unixland

Except '#' doesn't work, see:

https://secure.wikimedia.org/wikipedia/en/wiki/C#

The number sign (pound sign, octothorpe, etc) separates the fragment identifier from the URL. Browsers don't send characters after the # to the server. It works with Twitter because they use Javascript to AJAX in data from additional requests. http://en.wikipedia.org/wiki/Fragment_identifier

not really, the server can't see that # in the URL, it doesn't get sent by the browser. The only way they could do that is if they switched to #! style ajax urls ala twitter