[Javascript] Re: Charset (sort-of OFF_TOPIC for js)

[Scott Reynen]

On Aug 14, 2006, at 10:32 AM, tedd wrote:

> At 7:08 AM -0700 8/14/06, Bill Moseley wrote:

>> What should display for a byte with a value of 192?  You have to tell
>> the browser how to translate the string of bytes into characters --
>> and that's what the charset setting does.
>
> Okay, assuming your value is HEX, the code point 0192 is Latin  
> Capital Letter A with grave.

Documents don't exist in code points.  They exist in bytes.  A byte  
with a value of 192 only maps to a Latin Capital Letter A in ASCII.   
In UTF-8, it's the first half of a two-byte character.  That's a good  
example of why you should always state your character encoding.

> The point is that I  (perhaps I'm showing my ignorance here) don't  
> know of any characters in the Unicode dB that can't be shown by  
> UTF-8 -- so what's the point of other charset's? Are they for  
> legacy concerns?

There is a large and growing body of data that already exists in  
encodings other than UTF-8, so we can't just tell everyone to use  
UTF-8 and more than we can tell everyone to stop using COBOL.  There  
are also some characters that don't exist in Unicode.  See:

http://www.unicode.org/faq/han_cjk.html

> And, as I've asked before, does any of this affect javascript? Will  
> having the wrong charset cause js to fail in some way?

Character encoding affects any JavaScript working with server-side  
documents, e.g. JSP or AJAX.  If the server-side document has a  
different encoding than the client-side document, you'll need to  
convert before displaying.  But if you're not actually having that  
problem, I'm not sure why we're talking about this.

Peace,
Scott